Even if you did export your data, you can't export your friends to a competing social network. The real reason there's no threat to Facebook is because of the network effect.
The concept of network effect has been taught by B-school professors as a rule or law rather than a theory. Not enough research has been done to show how network effect can be broken up since it is just recently (last 10 years or so) been taught. 5 years ago, before Facebook started stealing MySpace’s users, you could have put the same wiki link and said the same thing.
One could even argue that Facebook is destroying their network effect by not allowing people to see all posts. The fact that Linkedin, Meetup, Twitter, heck even the new myspace all exist, there are lots of networks with varying degrees of network effects which could slowly take facebook share until the tipping point is crossed.
I’m not predicting facebook’s demise, rather I’m stating that network effects are important, but have not been proven to create a long term sustainable competitive advantage.
There are two barriers to entry: the network effect, s.t. the utility I derive corresponds to the friends I have on the social network, and their utility derives partly from me being on the social network, so at a critical mass it's hard to justify going onto another social network.
The network effect kicks in at a certain critical point. If you created a really compelling social network and added the feature s.t. people could say, "import from facebook", choosing your facebook login/pw, then if enough people did this, critical mass would be reached. The network effect just states that this critical mass might take awhile to reach, but it's not impossible. Facebook disallowing you from exporting your data makes it so competing networks cannot reach that critical point quickly enough.
http://en.wikipedia.org/wiki/Network_effect