You're going either too far or not far enough here. Chess has no correct set of moves, but rather a theoretical tree of optimal moves that is so enormously big that I will confidently claim that non-enhanced humans will never be able to memorize it. Either way, if you were to assume that humans were capable of memorizing it, you should also assume that they would be capable of making the exact right decision (based on probabilities) at any time in a game of Dominion, leaving the result up to chance alone. However, a game that's decided on chance alone isn't more than a coin flip, and therefore just as dull as a game with fixed winners.
The point is that Chess' actions depend exclusively on your opponent's actions, while Dominion's don't. You can like either and complain about either. I like both.
We have been down this road before.
For people who want to be super correct: chess is a low-luck game, but does have luck. This is trivially true due to the impossibility of complete analysis. As always I then recommend Richard Garfield's speech on luck vs. skill, which you can find on youtube.
I'm going to disagree on an abstract level, ignoring actually playing the game.
I'm all but certain chess will be brute-force solved in my lifetime. And even if it's not, we can show easily that such a solution exists even if computing technology as of 20xx or even 23xx can't find it. (The computers described in Star Trek almost certainly had solved chess.)
This is not true of Dominion and other games that involve randomness. Dominion, like poker, is non-solvable.
Chess has no such randomness; you cannot have a "bad beat" in chess, because there are only the moves that have been made. The same could be said for many other games: Go, Caylus, Prismata (though setup is randomized), Terra Mystica (random setup), and of course dozens of games that *have* been solved.