Mature Market leaders don't usually have the network effect lock-ins that Facebook has. In a traditional market, I'll swap to the new Product (Product X) as long as Product X does a decent job on the features I use in the original product and something new that makes it better for me.
The problem is that a Facebook-Killer has to be able to attract everyone at once, because of the network effects. It can't just be better than Facebook for me, it has to be better than Facebook for most of my friends (who use completely different subsets of the Facebook features), because I'm not switching unless they do. The end result is that the Facebook-Killer, in order to build its network effects, has to have replicated features and new features that make it better than Facebook for a large segment of the population.
A traditional app-killer can duplicate the most used 40-50% of the App's features and toss in a new revolutionary feature, and it'll take off (since most people don't use App's features that AppKiller didn't implement). However, Facebook is like Word - most people only use 20% of the features, but everyone uses a different 20%. So if you want to grab even a third of Facebook's users, you have to re-implement 60% or more of Facebook's features in addition to your killer features.