Wow was I tired last night. I'm surprised that post came out as readable as it did. I don't think I finished what I called the Outgroup problem. Lets try again.
Another problem - there is always an outgroup. The outgroup to your nuclear family is your cousin, the outgroup to humanity is the chimp, the outgroup within humanity I expect would be pacific islanders or those guys who made it all the way to the tip of South America. So you do tests of genetic similarity and it scales whatever level you examine it at. What criteria do you use to decide where to stop?
Its like human populations have some fractal-like properties. Imagine you've got a Google map of Humanity. Nah, lets go one better. Imagine you've got a google map of Life and that you have zoomed in far enough - You started at multicellular organisms and spun your mouse wheel while you went animals --> vertebrates --> mammals --> primates --> and further. You've reached the point where both chimps have split off from humans and then you tick a box to filter out species that didn't make it to the present day (because our job is hard enough).
Ok, so, we see one group representing all of humanity. Now lets zoom in further and get some more detail. Our magical app performs a statistically significant bunch of mathematical cluster analyses is real time and one group becomes two: Most Africans, and [Other Africans + Everyone Else]. Zoom one more click in and the three groups are Most Africans, [Other Africans + a mess that can euphemistically be called Indo-Europeans], and [East Asians + American Peoples]. If you zoom in to about 6ish/7ish groups then when you will probably see, depending on how the analysis goes this time and the genetic dataset, is Africans, Other Africans, Caucasoids, Southern Asians, East Asians, Americans, Pacific Islanders.
But there is no reason to stop there. Zoom in on any of this groups and it'll start splitting off subgroups. There is always an outgroup and you can keep zooming in until its you and that nasty cousin.
Take home message: We haven't yet agreed on which level of zoom is the "realest", whatever that means. There may not be one. Or it may be discovered that there is one that doctors find useful while anthropologists prefer another. We may not live long enough to find out.
As for the last problem I am Christmassing with my parents and attempting to find Speciation by Coyne and Orr which may or may not be somewhere in their attic. I'm not getting my hopes up.