AI research facility priority

Well, it could be that the situation simply exposed a corner case in the facility priority logic, because I was playing with no warp points (starting tech of 5 mill), and the AI refused to build research facilities unless the planet had less than 100% on all resources. This could actually be because, despite being isolated, they insisted on building up massive fleets, so it could be they simply needed to pay upkeep costs. I imagine that is a much more difficult problem, because it is not easy to decide if you should build a big fleet and have less research, or keep a small fleet and be vulnerable. Perhaps the AI should err on the side of one or the other depending on its characteristics (more scientific vs. more aggressive), especially considering how fast ships can be built in SEV.
While we're on the subject of the balance mod, there is another more disturbing issue I noticed. I noticed that when a lot of the AIs got their hands on ECMs, they would design fighters and troops without weapons (not an uncommon issue in stock too...). I guess this has something to do with component priorities, because they usually had full engines, combat sensor, ECM, small supply, and no gun. Maybe the small supply could be lowered in priority? And the combat sensor should probably have lower priority than guns, since it's useless without a gun... I personally never put supply on fighters or troops, because they hardly ever seem to run out, and the extra space could be used for armor or guns.
Also, I noticed that there is logic in the AI code for using stellar manipulation. Is the AI actually "fully certified" to do stellar manipulation now? I've never seen it do that, but it could be that the 5 mill points it started with simply wasn't enough. Is it smart enough to "explore" with new warp points if it's isolated? Because that would actually enable no-warp point maps to be fun in singleplayer.

Ah, OK, I see.
I thought I saw a post on the BM section of spaceempires.net suggesting that stellar manipulation was in.
I noticed the diversity in design types and weapons - I thought that was actually pretty cool... if only there was more component diversity in the game itself (because really, there's not much to vary besides weapons - everything else can only really be 'the best of'). It also seems (and I could be wrong in this) some races prefer different strategies - I've seen some build mines, while others focus on fighters, while others build lots of cruisers. It could have more to do with research priorities though.