In regards to dynamic vs static generation of good/bad:
Ahha, I hadn't realized that this was now the case, I've gotten to the point where I generally know what units are good against what units and don't need to use it for much. Also my test case the ai had so many different unit types by that point its was hard to tell
. I agree that turrets/starships need some looking at though.
In regards to the spire:
I had thought that the 2 main advantages the spire had over the zenith was a) higher boost range and b) ammo type(it can target things near ai home command centers/anti-missile turrets since its shell rather than missile). Not going to complain too much about a range boost though (though it will make taking out enemy home worlds tougher since they tend to have at least 2 or 3 spires under the shields defending the ai command center). On that note, I wouldn't mind seeing anti-missile turrets be slightly more common for the ai.
On the general topic:
Honestly I tend to not react too much to enemy ship concentrations. I'm with haagenti's comments in the attritioning the enemy post that I play against the meta game, and while I do react somewhat to the ai ship types(moreso early on), I often rely on numbers to make ships that are not good but not that bad work against units(300 cruisers will generally do pretty well against any 20 or so non-starship units even fighters) and focus on creating situations where I can bring overwhelming force against the enemy(and defeat it in detail).
My biggest problem with the good/bad numbers is the fact that they don't take into account range and while I do understand the reasoning for it, this makes snipers look useless and tends to overvalue short range/slow units. In reality if my cruisers/deflectors/etc. are going to be facing tanks(or some other slowish lowish range unit), I'm going to kite the tanks and take pretty much no losses, even if the tanks claim to do well against my cruisers(I don't remember if they do, just an example). Even against units that are faster than cruisers slightly, by moving directly away you can reduce their effective speed to ~4-8 for many units, giving you plenty of time to take advantage of the high range. On the flip side, I don't see any way to capture this better without biasing the data in the direction of the ranged units and making units good at wormhole assaults look worthless.
