In NPR's 13.7 blog, Ursula Goodenough writes of recent research on hominid genomes, and the increasingly varied story of our family tree. To take just one tidbit, it appears that the locals of Papua New Guinea are "roughly 92.5% African, 2.5% Neanderthal, and 5% Denisovan"--that is, this population (with which the rest of us are interfertile, of course) has genetic material belonging to two extinct hominid species, whom we had previously believed were distinct from(read: non-interbreeding with) H. sapiens. We knew we shared common ancestry with them, but the notion that they are among our direct ancestors (for some modern populations, at least) is new.
As I tell my students, Darwin's Origin of Species was ironically titled, because evolution by natural selection makes obsolete the concept of "species" as it was used at the time (or as creationists still use it). Rather, as populations vary across time and geography, black and white distinctions simply do not exist. Take the example of ring species, for instance, where neighboring populations can interbreed just fine, but populations a bit further apart (geographically, or chronologically, it works both ways) cannot, despite a continuous line of interbreeding populations linking the two. Where is the species line to be drawn? Are these one species, or two?
Frankly, it's a bit like Pluto. Pluto is what it is, whether it is called a planet or a planetoid; our linguistic handle on it is for our sake, and simply allows us to talk about it. The concept of "species" is a similar abstraction; tremendously useful in some cases, impossibly vague in others.