But that is exactly the point. A WW1 artillery piece will likely deal more collateral damage than a modern howitzer with GPS guided ammunition, even though the second will me much more lethal. Of course WW2 bombers inflict much more collateral damage than WW1 bombers, which has a lot to do with how much damage they inflict in the first place. (Well, collateral damage was pretty much the the usage of heavy bombers)
So overall I would say the category of weapon should be more important than the tech level of the weapon for how much collateral damage is inflicted.
Except none of that has anything to do with armour technology and everything to do with targeting technology and techniques. Being able to park a 500 kg bomb directly on top of an enemy fortification to destroy it means you only need 1 bomb, but if you've got an 80% chance to miss you are going to drop at least 5 and probably 10, or more, because then you can be very confident it's been destroyed.
If anything, this is an argument that who has the ECM/ECCM advantage should have an impact in collateral damage calculations rather than armour or weapons technology levels.
Yes, it is modified by tech. More powerful versions of the same weapon type will do more damage. As defenders improve their armour, it will take more firepower to overcome them.
Also, as tech levels rise, population and industry will likely be larger so the collateral damage may not be much different in proportion.
Well, given that the number of combat rounds and the number of shots fired per weapons category is likely to remain the same roughly speaking I'm not so certain about that. And there's a few units that get better armour relative to low level technology, so that's even more shots fired. If anything,
more collateral damage is likely to accumulate during a planetary invasion relative to the population and facility numbers rather than less or equal.
And planets with high fortification modifiers suffer more collateral damage because it doesn't care about fortification, but unit combat calculations do.
Damage to rubble doesn't cause civilian casualties or dust. The assumption is that civilians generally won't be in destroyed areas and the fires that would contribute to atmospheric dust have burned through already. It's not perfect, but it is intended to simulate that collateral damage has diminishing returns if an area has been fought over for a while.
This gets
really weird once you start ground combat on planets with no atmosphere then. I get your point that on planets with a breathable(-ish) atmosphere and close enough to their star to support growing crops with minimal infrastructure support (so basically very low colony cost planets) atmospheric soot caused by fires resulting from combat are a major concern, but a planet that's not capable of doing that and has no (native) biosphere will not have the sort of fires that impact atmospheric dust. Either there's no atmosphere, so any dust and soot enters a ballistic trajectory that's going to fall back down in days at the most and more likely seconds or minutes, or the atmosphere is so unhealthy to crops and the population in general that all biomass is internal in the colony's infrastructure and there's a not inconsiderable investment in the atmospheric processing equipment to scrub dust and other contaminants out of the air without it ever getting ejected into the atmosphere. And this gets worse if the planetary atmosphere pressure is higher than the native population's tolerances.
So... collateral damage should probably
always produce atmospheric dust, although certain classifications of planet may produce more/less dust than normal.
STO units can only be attacked once they fire.
There are two arguments here. One is that once they move the enemy can't tell what they are. On the other hand, once one side identifies a hostile formation, they will continue to know its capabilities. At some point ELINT will be extended to cover identification of hostile formations based long-term observation, in which case, the STO units will be flagged as such when their formation is identified even in the first scenario.
Right. That's fair. And I get that the intelligence side of the game needs a lot more work. It'd be nice if extended ELINT observation would eventually let you target enemy formations directly with orbital fire.