I guess an old but still worthwhile plug is Brads old thread.
What do the AI modders need (tools, etc.) from the devs in order to provide more access to AI modding?
What do we want to see in terms of AI behaviour going forward, and hence, what do we hope that Crusade will improve upon in terms of AI?
I did only do a quick check on item 1, which lead me to conclude it is not addressed yet. Item 1 is probably the largest bottleneck for the AI on all dfficulties, because it occurs quite early and is really hard to compensate via the accessible .xmls. But build-orders determine a players performance until late in the game and the AI's build orders are weak early on. Some of it can somewhat be fixed via .xml allready, but item 1 I got stuck on a few months ago.
That's all I can say so far though, I was a little preoccupied and didn't test 2.0, yet :/
I could imagine SD ignoring it in order not to mess with beginner and easy difficulties (see Brads point one about AI expectations). And because they prefer to 'fix' it with cheaper handicaps. I dislike the handicaps because they kinda remove personality as compared to better builds, which will specialize the AI in some way (game balance permitting), opening weaknesses as trade-offs, too.
There is also Horemvores wishlist, which has a few AI items.
How can the community help make improvements to the AI less of a thankless task in the future?
I am tempted to advice expanding the modders interfaces. If the game remains popular, some players will inevitably surpass the devs understanding of the balance, especially if multiplayer takes off. Those will most likely be able to build the best AI. Unless payed for, AI improvements will remain a thankless job to SD, being object of the disappointment of both: the people for whom the game is now too hard and the people for whom it is still not hard enough.
What may be an interesting idea, is taking a leaf out of blizzard's book and occasionally featuring mods, incentivicing feedback to the modders.
Last but not least:
Don't be like me suggesting solutions/implementations x). Without knowing the sourcecode the solutions are almost allways simpler in your mind than they are for the actual devs. Just accept it. It's been my experience comparing bug reports I made for a starcraft2 map and having worked on it some time later. And the sc2 editor is still somewhere in between a hobbyist and a professional tool, I think, so the divide is probably even bigger.
From Brads old thread:
## How you can help ##
Helpful: Telling me dumb things you saw the AI do that made you able to beat it is helpful. Very helpful in fact. Any mistake that would allow you to win regardless of the handicap.
Not helpful: Things that could be addressed just as easily with a handicap. The AI doesn't build up its planets well. This is on our list. But it's actually not that important since that can be addressed with handicaps.
Really helpful: Saved LATE games with descriptions of what you are seeing. Maybe the AI isn't defending its planets well. Maybe it's sending out crappy fleets. Maybe its fallen way behind militarily. Maybe its ships are crappy. These are things that I'm interested in.