Isn't this a philosophical issue versus a practical one? I mean, yes, humans do it all the time, and they're wrong pretty often, too. I'd value computers getting to be as good as we are (heck, I suspect if we can get them that far, we can get them better than us) but I don't disagree with the philosophical absolute statement that data can't tell us about causes.

(Written without having read the article; I may come back and backpedal severely after I've done so.)