I suspect that one of the reasons that Facebook has released the platform API is to allow small teams of external people to their R&D for them.
There's probably a million features that might or moght not appeal to the Facebook userbase and they can't try them all out by themselves.
There's always the danger that they might simply rengineer any successful service but my hunch is this that they would be likeley to acquire any such site becuase they wouldn't want to discourage people building on the Facebook platform.
They wouldn't want to kill the goose that laid the golden egg, would they ?
Anyway, it's just a thought ...
That's not to say it won't happen...but I suspect it'll be even odds for quite some time that Facebook will develop their own version of the most profitable niches (already happened with classifieds, apparently, which was announced just days before another product from another company that did the same thing). I don't follow Facebook very closely, so I don't know any details, but if you build a company based on Facebook, the only liquidity option is "sell to Facebook". I don't like those odds.
In other words, if you want to work for Facebook, just get a job with Facebook. There's still tons of upside on their stock, so if you get options, you will likely be rich in a four years when they vest.