I have no inside knowledge and I speak only as a research student working on search engines, but my guess would be that Facebook and Twitter don't allow their pages to be crawled[1]. If I recall MS had to strike a deal with both of them so that Bing could use them in their results. Another "problem" is that the notion of document relevance varies from a "standard" web page to a social network[2], and mixing them up on the same result page could be confusing.
[1]: a good social search would need realtime access to the API of those websites, which is why you still see "some" content like FB profiles in Google's results.
[2]: what I mean is that you have additional parameters to take into account: incoming/outcoming edges, "status" of the author, additional metrics of popularity, redundancy of content... Social search is an entire research topic of its own.
