Refuelled, following up on this.
Looking again at the patch notes you posted, looks like the idea at that time was basically option "Ignore Strength Completely!" (I'm going to edit that prior post and reorganize that list a bit to make it better, people are going to get confused...)
Thinking about it, I can see why it changed from that earlier idea somewhere along the way -- taking strength into account prevents cheesy stuff like cramming ATs full of noncombatant units like Engis and Scouts from producing an armed to the teeth AT.
On the other hand, even if you were to cram every zero-damage unit you could make into ATs, you'd only fill what, one? One and a half? With hacker products no longer transportable, there really isn't anything you can infinitely spam to stuff ATs with now. So this way of doing things might now be doing more harm than good in normal play.
On the other, other hand, taking strength into account makes the mechanics consistent between the AI and the player. All combat-capable container units calculate their damage in the same way, which is generally a good thing from an intuitive play perspective.
But, counterpoint, the player and the AI don't tend to build fleets quite the same way -- you have to run the AIP into the stratosphere to see a single AI fleet on the board that's pushing 30-50K strength, at least at difficulty 8 and under, but I get the impression most of the player base plays difficulties in that ballpark so they should probably be what guides balancing this the most.
I didn't see cohesive AI fleetgroups over 30K until I had AIP over 650-700 points. The AI carriers are also slow and non-cloaking, they pass damage directly through to the containing units unlike human ATs, and they tend to start unloading them quickly rather than using them as-is until they pop, then turning the fleet inside loose to wreak havoc. So, there's a valid argument that ATs and Carriers are NOT analogous, and thus *should* have different rules for computing damage.
Writing this out, I lean towards "ATs and Carriers should be handled differently" -- mostly. I'd favor going with either the hard cap or soft cap on AT strength approach, and leave Carriers as they are so they keep their nasty alpha strike as compensation for their slow speed and inability to protect their contents. They lack defense/utility, so they should keep their DPS. ATs get a ton of defense and utility, so they should have their damage cut back.
Out of both, I'd prefer the soft cap, partially because it would be easy to write, and because it steps down the damage under all circumstances, so there's no weird "little power --> TONS OF POWER --> 3x as much power inside doesn't give you any more power over the previous step". Brickwall caps like that are always a little bleh IMO, very brute-force. It maintains a smooth climb in DPS gains as your fleet grows in power, producing deadly ATs only in late game, which is when the AI is going to be breaking out the big guns to match.
This change may possibly affect AI Barracks, but I can either insert a chunk to handle them the old way, or just let it be, since Barracks protect their units like ATs do, so they're getting an extra benefit Carriers don't and thus have compensation for losing some DPS. Besides, half the time it's irrelevant anyway -- the Barracks will pop when the command station goes, so if you're capping the planet you don't even have to bother with the bunker, let it unload and smash what comes out. Not something I'd lose sleep over.
