Here's another few I'd found previously on the topic of faceted identities; There's been a bit of research on this topic in the social sciences for a while now, for whatever it's worth, but here are some papers that I found insightful:
Danah Boyd's undergraduate thesis (Danah is now a Sr. Researcher at Microsoft):
http://smg.media.mit.edu/people/danah/thesis/danahThesis.pdf
A more recent paper from Yahoo: http://research.yahoo.net/files/pr308-farnham-1.pdf
And this much older paper makes it clear that the idea of personal identity being multifaceted is all but a given in many lines of research, pointing back to influential papers as old as 1968: http://www.ssc.wisc.edu/~jpiliavi/965/hoggterrywhite.pdf
So, honestly, I've always been a little surprised that supposedly smart companies/developers at some of the hottest tech companies (FB & Google, for example) would build business models/big plans around such a simple model of social identity when it's so at odds with the way people really work.
http://blog.pinboard.in/2011/11/the_social_graph_is_neither/
And yeah, I have two Twitter accounts: one for friends and one for business.