You're leaving out the pesky detail that an entirely different experiment with an entirely different sample using an entirely different communication method yielded a very similar result. This is damn interesting. Not conclusive by any means, but interesting.
A sample of people from Nebraska and Boston. Hardly any better from a global perspective (indeed, probably worse, since MSN is fairly universal). What about the 80% of people living in the developing world?
Consider that they're linked through people like me. I'm personally currently linked to people born in ten countries off the top of my head, and have previous links (former coworkers and acquaintances) to at least ten more. I am two links away to someone in every middle eastern country, nearly every other Asian country, a few South American countries, and a smattering of Caribbean, African and European countries.
(I am American, but have a few friends who came to the US to study and work.)
You're right, "entirely different sample" is a bit of an exaggeration. Assuming they've stumbled upon a constant for social interaction, it may be for our society rather than our species. There's no telling without further testing. But that the experiment using instant messaging didn't a have a significantly lower average degree-of-separation really does surprise me.
