You've probably seen the Atlantic piece about because by now. I actually have been wondering about this for a long time, mostly because I read Wonkette, rightly cited as a place where you can find this construction all over the place. And it's been talked about for a while.
What's most striking about it is not that it's a preposition — it has a pretty classic distribution of a preposition (followed by nouns, etc.) and the semantics are not weird for a preposition (German wegen means something pretty similar at some level, 'on account of') — but that it takes these bare nouns. Anyway, here are some things I've wondered about here … any insights would be appreciated.
First, the creation of a new preposition from an old clause-linking element / conjunction doesn't seem so weird intuitively but the direction of such change seems to usually go in the other direction … adpositions become conjunctions. (Harris & Campbell's Historical Syntax talks about several examples, pp. 291-293 and it's pretty common in the grammaticalization lit, I think.) Is this direction — conjunction to preposition — particularly unusual? Isn't a preposition 'more grammatical' in some sense than a conjunction?
Second, the history of because is kinda cool: The usual story (I don't have particular reason to doubt it) is that it comes from a prepositional phrase itself, by cause. There's a lot of work out there on cycles of linguistic change this days and this seems to echo those kinds of patterns, in the sense of Elly van Gelderen's The Linguistic Cycle. Is this a known cycle? (I don't recall it from that book.)
Third, are there other prepositions that can take the pattern of stuff this does? 'For' in a certain sense does only for mass nouns, maybe: She did it for family, Homer did it for beer, they did it for love. But you can't do something for desk or howitzer, etc. There's some semantic thing going on here, surely, right? The Language Log post on this from way back touches on the issue but there's gotta be more.