To players who haven't beaten Petra and don't want to learn how they work I've spoiler some bits of info. So I've checked and noted the AI's limitation and I'll be using pop per minute[PPM] as the term: VE will have max pop(300) at min 60[5PPM], E - 45[6.7PPM], M - 35[8.5], H - 30[10]. VH - 25[12]. New players will be ending the game around 1 hour mark, so [5PPM] is quite simple albeit if they can survive, tbh let it stay limited. - The AI on Medium and above will dedicate so there is quite a lot of idle time when they could use it to gather more resources, I understand its use for harrassing, I will still beat them, they will have to rebuild it = stunted economy. If they could transition from gatherer to non gatherer in late game then it will be beneficial. - In terms of Very Hard from what I observe when they solo, its possible they could reach 15PPM which is my current level(Max pop at min 18-23), 15PPM for VH will be an actual pain to fight but its what I recommend because its meant to be difficult but possible to win. So players who intend to fight 2 VH[26PPM], will have to unfortunately fight the improved [30PPM] lol.

Making "goodness" quantifiable is a good input. This way AI devs know what to strife for. As I understand the hard AI should produce 10 pop per minute (or ppm to introduce a unit ) What ppm should the very easy and very hard AIs have? Is pop per minute a good measurement? Maybe points per minute is better.

"However, artillery being captured, turned around in the heat of battle, and starting firing on the troops of its previous owners? Unlikely." That's probably right (can't remember any counterexample), but the same argument could be applied to ships, and that's going to be in the game now. It’s like proposing removing the building of towers because that didn’t happen while a battle was taking place meters away. It's all a representation of things, numbers and time are shrunk with respect to what in reality would have been, it seems to me that completely removing something that happened just because it didn’t happen in the same battle is the wrong call. Capture times would have kind of prevented turning things around in the heat of battle, and if not, a “use cooldown” could be implemented, one being able just to move the piece in the meantime. Edit: I think something like that was proposed for towers and fortresses, to avoid turning them around fast, although more from a gameplay perspective than a realistic one. For ships it would also make sense to have a use cooldown.

Here the opinion of Nescio:  

If one's civilisation has the arsenal, then it should be assumed that there would be engineers in the army that know how to handle equipment, it makes no sense to remove something that happened in reality because there’s not a differentiated engineer unit in the game. I’d even argue that even some things like rams could have been used by anyone. Without the arsenal maybe one could capture and move, but not use (the Rhodians captured very complex siege equipment from the Macedonians, which they sold to finance the Colossus). In reality equipment was even reverse engineered (the Romans used Carthaginian techniques after capturing one of their ships), but this would be too much, so what I mentioned before (having the arsenal) should be a reasonable minimum requirement for using captured equipment, without altering what one’s civilisation can build (which is harder than just using), to make things simple.