A Potentially Informative Experiment

Suppose one were to start with a game between any two Grand Masters.  Any game that did not result in a draw, let us say.  The game would not have to have been played here.  Anyway, then get the best minds from Chess.com to determine where the game was lost.  What was the first mistake made by whichever player lost?  What would have been a better move?

Now go back and make this better move, say in a vote chess match between two groups of top players here on Chess.com.  These two groups keep playing, with the provision that at any point, either team can take back moves, and replace them with what appear, in hindsight, to be better moves.  The point of this exercise would not be to win, or to determine which group is stronger.  In fact, it could even be the same group making moves for both sides.  No, the goal would be to converge to the perfect game.

It would seem that this approach would have a fair chance to lead to the winning stratgegy in a two-player game with no ties.  But in chess, I fear that after a while both sides might get stuck in a quagmire that ends in a draw.

Still, I ask, is it worth trying to do this?  I mean, in theory, any idiot could play like a master if allowed to take moves back and replace them with better moves, right?  So imagine what could be accomplished if masters were given the benefit of this flexibility.