I think being difficult "for computers" is a misleading concept. Go is more difficult for our current algorithms than chess, at least in part because we've focused on chess for much longer.

Sooner or later we may discover some new algorithm (this deep-neural-net-based approach may be it) which will reduce Go to be as tractable as chess, or even more so.

It was more difficult for computer programmers to program an ai which reach a human top level, and for this level you have to throw much more hardware in it.

Until it isn't. Once humans are handily beaten in both games, it's kinda hard to tell which game is harder (I guess you could compare the amount of computations needed to beat the top human player, but that does not let you distinguish if it's harder for computers or conversely easier for humans.)