> Have we ever really defined species barriers?

It's fairly easy to make definitions, and there are several. The real problem is that many biologists for many decades have been confused about whether we are attempting to make pragmatic definitions or whether we are uncovering "true answers" regarding biological discontinuities. It might not seem that bad if you don't consider geographic separation, but when you do, the literature turns into a total mess. The truth is, though it's unpalatable to many, that there's nothing about biological science that implies that the question "are these two geographically disjunct populations members of the same species?" has any particular answer.