The problem with that line of thinking is that it requires outside knowledge to distinguish. True, not everything is inherently broken, but more often than not, it's easier to blame the tool than to blame the user, because more often than not, it is indeed the tool that is broken.