Okay, I suggested adding artifacts/techs that affect espionage, and Andrew pointed out that the espionage subgame is pretty skewed in favor of the player already, and beefing espionage up would probably just make the AI easier to take advantage of. This is probably true, and adding complexity for complexity's sake is poor design. It did get me thinking about ways to enhance the game that would improve, instead of detract from, game balance, though.
I remembered an old article on game balance written by Brian Reynolds, where he talked about the "Poor get richer" rubber-band effect. The article is here:
There is a definite tendency in STUN (and likely in STUN2, which I've yet to put as many hours into) to just steamroll over the AI once a lead is gained. What Reynolds proposes, and what makes the most sense for gripping gameplay, is that the further ahead you are, the harder it should get to further increase your lead. This principle may well be in STUN somewhere, but I don't know anything about its internals. I would suggest that if it's there, it's not as pronounced as it could (should?) be.
To this end, I would like to propose, and encourage further proposals/refinements of, ideas that would employ this principle to make STUN2 more of an edge-of-your-seat kind of experience.
The espionage tweak to disallow stealing more wisdom artifacts when you already have 4 was a step in this direction, albeit somewhat artificial. What if it was just exceedingly difficult (but still possible) to steal wisdom artifacts from races that only have a few, the rationale being that because they have so few, they'd be more heavily guarded? Conversely, if I've got 8 artifacts, I'm more likely to lose them through espionage.
As far as adding espionage artifacts/techs (Master spy, Personal cloak, etc), perhaps making the AI more aggressive in its espionage efforts would balance them out?
I'm also thinking there isn't enough of a risk in spy missions...I'd like a "critical miss" possibility, say, all your spies and informants on that planet get exposed (Spy was caught and tortured and gave up other spies). As well, espionage blunders should increase unrest in a publically embarrasing Hey-we-captured-Gary-Powers kind of way. Which brings me to unrest...
The philosophy-flip when unrest is high is a good effect. I would also like to see defections/revolts. A gunboat could defect to an opposing side or just become an independent pirate. A city/planet could rebel, requiring assault pods to recapture it. The empire should show signs of stress before abandoning its philosophy.
To bring this back in line with the "rubber band" principle, and for this to have much of an effect on the player, the AI would have to more aggressively promote unrest. Maybe unrest should increase "naturally" as an empire grows/population increases (anti-globalization protests/overcrowding). City facilities that help pacify the populace could be added as a counterbalance. Small empires should be more resistant to unrest (smaller population == easier to control).
Now, random events. All players could have a chance to get positive/negative events, but leading players could have a greater chance to get negative events, and vice-versa. Some ideas (shamelessly stolen from MOOx):
New mineral vein found on resource world
Merchant donates money to empire
Resource world dries up
Market crashes, treasury depleted
Diplomatic disaster (assassination attempt, etc)
Hm...I've rambled enough. What does everyone else think?