In the last month, I've won in every way (with each race-type I could think of), and while the conquest is perhaps too easy (a 12,060 point win on challenging with only melee builds highlights that), there are other situations where the other victories are viable alternatives. One of my first games was against a powerful Empire (I spent too much time exploring, rather than expanding), so I allied with the other Kingdom and used the diplomacy screen to pay him (lots) to go to war with the Empire. Then I waited about 5 turns and piled in through another front. That was a satisfying Alliance victory--while the particular setup doesn't happen very often, it's hard to see how that can be improved. Ideas?
In one of them, I played Pariden and set my sights on the ultimate magic victory. If you make that decision early on and push for it, you don't have to be dominant to get there (only the most developed cities--doesn't mean you can match the hordes of enemy armies). Not sure there is any way to improve the mechanics of this victory.
It stands to reason that if you build ridiculous amounts of strength every time, that can be considered the generic path to victory, and the specific avenue is moot. That's different than saying it's necessary. As release nears, I hope we see AI that will try to prevent the magic and Quest victories as you get far along the path. (The master quest is lots of fun, and that's really all that should matter--who cares, really, if you could have won in another way?)
In general, though, most situations/play styles call for an area-wide beatdown. Nothing wrong with that, but having those other paths available is more important than trying to tailor must-do-to-win paths for each of them, right?
And Kongdej, if the AI was any less aggressive, the game would be so boring it'd be unplayable (surely, most people are similar to me in that they hate it when the computer opponent simply waits around to get beat, all passive-like).