I don't really see how you can build an anthropic argument out of that, though. The idea that if you make a radical mutation in one aspect of a life-supporting universe, then it no longer supports life is probably not particularly unusual. For example, if you make the game of life 3D using the same totalistic rule then it no longer supports life either. That is just a consequence of dead universes being more common than living ones, and doesn't have anything to do with there being something special about the dimensionality of our space-time.

And cellular automata don't select for intelligence, so it is at least reasonable to conjecture that most observers evolve under physical laws of character similar to ours (and therefore, by the orbit stability argument, in three dimensions of space).