In response to an argument that agnostics are intellectually dishonest, Clark Goble at Mormon Metaphysics suggests that perhaps the difference depends on how one conducts the inquiry, and the atheist who insists that it can only be rational if it's in the way that leads to atheism is perhaps wrong in assuming that an honest inquiry must go exactly that way. I think the same can be said for how theists might think about atheists and agnostics who claim that the evidence leads to disbelief in God, and I'm sure Clark would agree.
Blake makes an interesting observation in the comments that what counts as publicly verifiable (or, I would assume, falsifiable) based on evidence seems to vary from person to person because people have different accounts of what is good (and thus of what a good God would allow). Is it verifiable that some state of affairs is better than another or that a certain conception of omnibenevolence is the right one? It assumes we can verify an account of the good. We can't treat such things as if they're of the sort that evidence speaks well to, at least in a publicly verifiable way. (The guy he's responding to nicely illustrates this, in fact, in his subsequent comments, by his assumptions of what counts as good or best.)