So, currently, combat has the following points of note.
Characters can seek cover and fire from it.
Weapons have varying ranges at which they are functional.
Weapons have varying damages.
Turrets have positions they are at.
Turrets explode when killed.
Walls can be placed or removed funneling enemies.
Walls can be destroyed by enemies, and will be if they are man-made.
A fairly straight-forward list. The implications of it mostly lead to something akin to a tower defense mode. Defense is ideally luring you enemy to enter through a kill-zone. Weapons are best used at range from cover to kill enemies before they can near. Bunkering in and having your walls attacked is the worst sort of thing that can happen, because fire is super-deadly.
So then, how do you make defense more interesting? How do you avoid it from becoming a 'tower defense' game, and how do you lead to have interesting implications?
1: Enemy assault logic, walls, and caves.
Enemy Logic
Enemies attempt to find an easy way in, even if that leads to their death.
Enemies will burn down walls to get in if this is a pain.
Enemies have no discernible goals and are just there to kill all targets.
Problems Here:
This means enemies can just be funneled into their death.
Interesting problems from enemies diminish.
Caves:
Cave walls are indestructible by enemies, thereby becoming the ultimate way to funnel enemies into an area.
Cave areas are markedly easier to make than normal walls.
Problems:
These areas are just great.
They have no faults.
Walls:
Walls can be made cheaply and are necessary to define outdoor areas as indoors.
Walls resist damage, but not fire effectively, making them poor as base defenses.
Problems:
Because walls are easy to kill and a poor decision, they ultimately are only used if the enemy is just not going to attack them. They create a sort of cheap 'building stuff' and no longer create an interesting framework of interactions.
Base Design:
Buildings need to be indoors to not suffer major issues.
Buildings can be a functionally self-contained complex, and their wiring means that they have major issues creating more 'sprawling locations'
Functional base design also means that choke points to ward off raiders are very important.
Problems:
These complexes have no need to see the outside, often ever.
Base design is very limited by the need to ward off attackers.
There is no reason to desire outdoor (non-mountain) locations.
Solutions:
Raider Logic: Make raiders retreat from heavy fire. Give raiders a goal.
Raiders should be
1: Limited by food, so they can be 'starved away'. This provides non-standard solutions.
2: Sufficiently reasonable that they will run away from a gun fight they are losing.
This means they will try to burn down walls if there is no other way in and this was a death trap. This means they can just try to recalculate their pathing at decisive moment in combat as though this region (between them and the attacker) is not acceptable. This means they will likely have to burn down walls to get in to a well defended location.
3: Interested in 'things' besides murder.
Being able to 'steal' resources from stockpiles (and they will steal as much as they can carry).
Trying to incapacitate a single (or more at higher numbers) citizen to sell to slavers.
Trying to set up in your base as home, without your leadership (kill all, with building destruction deprioritized).
These add character to the assaults and mean that certain methods of defending are unacceptable/insufficient.
Overall, the avoiding deathtraps is likely enough to make things a bit more interesting here.
Caves and Walls: Grant gameplay for each, make raiders interact with caves more interestingly.
1: In order to limit caves as a defensive mechanism, I recommend granting raiders the ability to bypass them. This can be done with 'high explosives'. Some raiders will randomly spawn with these devices, and will place them at locations where blowing up the cave wall or normal wall will notably increase the speed 1.5x at which they can get to their target (whether that is a person or stockpile).
2: In order to replace these regions, and make new barriers that create a more interesting definition of 'base' structure, I recommend the ability to make expensive (10+ metal) walls that do not conduct electricity, are nearly indestructible (high explosives only, including your own), cannot burn, and cannot be sold. They will allow players to make more meaningful long term decisions about base design, while remaining sufficiently impractical and expensive that they are not the go-to material.
3: Walls remain as a cheap base making material that also conducts electricity.
Base Design: Reasons for openness, issues with self-sufficiency
1: Include travelers more meaningfully, likely as additional traders. Them needing to enter your base (especially if they need to bring a large vehicle with them) increases need for openness in base design and interesting interactions with the outside world.
2: More internal threats (like things that rise from the ground, or travelers that turn out to be traitors), also help influence a more interesting set of stories and base design. Having people break down and attack allies due to 90+% fear also might work here.
3: A gradually increasing need for people to go outside helps keep up interaction (-5 happiness per day they haven't seen the open sky).
4: Desirable elements from outside agriculture to research also help here.
Combat: Increase variation in combat. Re-balance attrition and desirable weaponry.
1: Better close range weapons would be a large step in making combat optimization more interesting. As it stands weapons just sorta become better in every way, including range and damage. Making a deadly short range shotgun, and a deadly melee weapon would be a marked way to improve things.
2: Permanent health damage without medical supplies. Weapons are not random (not very random, at least), so you can count on everyone being able to do their part by soaking a few bullets. This isn't terribly interesting for gameplay, and making it cost a resource to take damage (medical supplies) is likely a good way to balance this out. It helps make the situation more dynamic over a longer period, without just relying on 'more'. This is probably best achieved by allowing characters to only heal a fraction (75%) of the damage they take, without medical supplies. Enough that the attrition won't get them, not soon at least.
3: Morale in combat. Fighting is terrifying. Having it constantly build up fear, then hit you as a panic attack during combat creates an interesting reason to have happiness instead of fear, and means that you can simulate the enemy panicking (thanks to your giblet cages) by the same system. This means you can have an interesting choice between easy loyalty and more competent soldiers.
4: Degradation of guns. Having guns degrade with use allows you to simulate ammo effectively, and gives you something to do with all of these spare guns. Creating a new class of 'gun part' (high tech components) allows you to abstract this without too much trouble.
5: Better enemies not more enemies. By allowing ranged to decrease your chance of being hit with ranged attacks (slightly) and melee to decrease your chance of being hit with melee attacks (notably), you can have enemies that continue to scale, without just giving the player better equipment. This becomes less important if 4 is done.
Characters can seek cover and fire from it.
Weapons have varying ranges at which they are functional.
Weapons have varying damages.
Turrets have positions they are at.
Turrets explode when killed.
Walls can be placed or removed funneling enemies.
Walls can be destroyed by enemies, and will be if they are man-made.
A fairly straight-forward list. The implications of it mostly lead to something akin to a tower defense mode. Defense is ideally luring you enemy to enter through a kill-zone. Weapons are best used at range from cover to kill enemies before they can near. Bunkering in and having your walls attacked is the worst sort of thing that can happen, because fire is super-deadly.
So then, how do you make defense more interesting? How do you avoid it from becoming a 'tower defense' game, and how do you lead to have interesting implications?
1: Enemy assault logic, walls, and caves.
Enemy Logic
Enemies attempt to find an easy way in, even if that leads to their death.
Enemies will burn down walls to get in if this is a pain.
Enemies have no discernible goals and are just there to kill all targets.
Problems Here:
This means enemies can just be funneled into their death.
Interesting problems from enemies diminish.
Caves:
Cave walls are indestructible by enemies, thereby becoming the ultimate way to funnel enemies into an area.
Cave areas are markedly easier to make than normal walls.
Problems:
These areas are just great.
They have no faults.
Walls:
Walls can be made cheaply and are necessary to define outdoor areas as indoors.
Walls resist damage, but not fire effectively, making them poor as base defenses.
Problems:
Because walls are easy to kill and a poor decision, they ultimately are only used if the enemy is just not going to attack them. They create a sort of cheap 'building stuff' and no longer create an interesting framework of interactions.
Base Design:
Buildings need to be indoors to not suffer major issues.
Buildings can be a functionally self-contained complex, and their wiring means that they have major issues creating more 'sprawling locations'
Functional base design also means that choke points to ward off raiders are very important.
Problems:
These complexes have no need to see the outside, often ever.
Base design is very limited by the need to ward off attackers.
There is no reason to desire outdoor (non-mountain) locations.
Solutions:
Raider Logic: Make raiders retreat from heavy fire. Give raiders a goal.
Raiders should be
1: Limited by food, so they can be 'starved away'. This provides non-standard solutions.
2: Sufficiently reasonable that they will run away from a gun fight they are losing.
This means they will try to burn down walls if there is no other way in and this was a death trap. This means they can just try to recalculate their pathing at decisive moment in combat as though this region (between them and the attacker) is not acceptable. This means they will likely have to burn down walls to get in to a well defended location.
3: Interested in 'things' besides murder.
Being able to 'steal' resources from stockpiles (and they will steal as much as they can carry).
Trying to incapacitate a single (or more at higher numbers) citizen to sell to slavers.
Trying to set up in your base as home, without your leadership (kill all, with building destruction deprioritized).
These add character to the assaults and mean that certain methods of defending are unacceptable/insufficient.
Overall, the avoiding deathtraps is likely enough to make things a bit more interesting here.
Caves and Walls: Grant gameplay for each, make raiders interact with caves more interestingly.
1: In order to limit caves as a defensive mechanism, I recommend granting raiders the ability to bypass them. This can be done with 'high explosives'. Some raiders will randomly spawn with these devices, and will place them at locations where blowing up the cave wall or normal wall will notably increase the speed 1.5x at which they can get to their target (whether that is a person or stockpile).
2: In order to replace these regions, and make new barriers that create a more interesting definition of 'base' structure, I recommend the ability to make expensive (10+ metal) walls that do not conduct electricity, are nearly indestructible (high explosives only, including your own), cannot burn, and cannot be sold. They will allow players to make more meaningful long term decisions about base design, while remaining sufficiently impractical and expensive that they are not the go-to material.
3: Walls remain as a cheap base making material that also conducts electricity.
Base Design: Reasons for openness, issues with self-sufficiency
1: Include travelers more meaningfully, likely as additional traders. Them needing to enter your base (especially if they need to bring a large vehicle with them) increases need for openness in base design and interesting interactions with the outside world.
2: More internal threats (like things that rise from the ground, or travelers that turn out to be traitors), also help influence a more interesting set of stories and base design. Having people break down and attack allies due to 90+% fear also might work here.
3: A gradually increasing need for people to go outside helps keep up interaction (-5 happiness per day they haven't seen the open sky).
4: Desirable elements from outside agriculture to research also help here.
Combat: Increase variation in combat. Re-balance attrition and desirable weaponry.
1: Better close range weapons would be a large step in making combat optimization more interesting. As it stands weapons just sorta become better in every way, including range and damage. Making a deadly short range shotgun, and a deadly melee weapon would be a marked way to improve things.
2: Permanent health damage without medical supplies. Weapons are not random (not very random, at least), so you can count on everyone being able to do their part by soaking a few bullets. This isn't terribly interesting for gameplay, and making it cost a resource to take damage (medical supplies) is likely a good way to balance this out. It helps make the situation more dynamic over a longer period, without just relying on 'more'. This is probably best achieved by allowing characters to only heal a fraction (75%) of the damage they take, without medical supplies. Enough that the attrition won't get them, not soon at least.
3: Morale in combat. Fighting is terrifying. Having it constantly build up fear, then hit you as a panic attack during combat creates an interesting reason to have happiness instead of fear, and means that you can simulate the enemy panicking (thanks to your giblet cages) by the same system. This means you can have an interesting choice between easy loyalty and more competent soldiers.
4: Degradation of guns. Having guns degrade with use allows you to simulate ammo effectively, and gives you something to do with all of these spare guns. Creating a new class of 'gun part' (high tech components) allows you to abstract this without too much trouble.
5: Better enemies not more enemies. By allowing ranged to decrease your chance of being hit with ranged attacks (slightly) and melee to decrease your chance of being hit with melee attacks (notably), you can have enemies that continue to scale, without just giving the player better equipment. This becomes less important if 4 is done.