I would have loved to have summed it up rather simply as that Facebook is story of accidental success. Zuck started the product to have some fun, and incredibly, five years down the line, the fun has not ended. But, it is not all fun and games and even with its astounding growth, the fact is that Facebook is struggling to find a clear direction in which it should head and it also having trouble finding enough revenues to offset its ever-increasing burn rate.
Facebook has zero value when it is not public: This makes for a world of a difference when you are trying to introduce a set of services on the platform that goes beyond its primary use case. Platforms that are primarily meant for public (restricted, yes, but public still) consumption never fare too well in the private world. This is the reason why you don't see social networking in your email inbox (sorry, Xobini is NOT social networking) and the reason why your online banking service does not have a 'social' aspect to it. The simple reason is that it is counterintuitive.
Facebook's main use is for people to be able to connect with each other, in an environment that allows you to pick and choose the audience. You can use Facebook without any connections, but that would also send the value derived out of using the product spiraling downward. In effect, the main use case for Facebook and the extended world it is trying to embrace work at cross purposes. Facebook is a network that functions on the basis of exclusion, while the newer things it is attempting works best on the basis of inclusion.
Mixing up identity with connections: Mixing your identity with the connections you make is a fairly subtle but significant mistake that all of us make. Who I am on Facebook is not my identity (though, poking thrice a day can say a lot about who you are), but it certainly gives me a dashboard on the state of my personal connections. With the new features, Facebook is trying to sew together my identity with my Facebook account. Other than the earlier mentioned counterintuitive problem, it also has a problem of profiling based on a restricted context which will always give you an inaccurate picture.
Context, not connections, drive content discovery: One of the major thrust areas for Facebook is peer-to-peer discovery of almost everything, which conveniently ignores some major points. 1) Friends are not always your peers. Friends are sometimes just friends. 2) The converse also is true in the case that your peers are not always your friends. Content discovery is almost always best done when it led by context (event, topic, theme), than by relationships (remember email forwards?). This is why half of the suggested content you see from your friends on Facebook is junk. You have much better luck with content discovery on Delicious, Digg and Reddit.
Context is often required to be non-shared and private: One of the reasons why you freely consume so much of content on the internet is because the context in which you would have found them are not public. If your Google search history could have been shared on Facebook it would have an immediate effect on what and why you would search for things. Man is often known for the thoughts he expresses, not for how the thought was formed.
The 80:20 art of network maintenance: Every network has a crucial dipping point where the effort that you put into maintaining your network starts to become more than the value you derive out of being a part of the network. After a point, it is hard work to have a clear idea of who gets to see what you do on Facebook and maintaining that alone is a huge chore. On a platform like Google, even if you don't share there is no core value erosion, on Facebook if you don't share or participate the network immediately becomes useless to you and vice versa.
This is the reason why I have held that social networks have an average of three years being at the peak, by that time everyone you know and their dog is on the same network and the last thing you need is to have parallel virtual universe that reflects the real life network you are part of. Basically, this is the point where the whole network effect starts to break down. And this is what eventually marks the death of every social network.
First, just be a good social network: Yes, it is a great thing to have BHAG, but that is only possible when you have done well in your primary context. Facebook needs to nail the primary need of being an easy-to-use and uncomplicated social network first before it gets into anything. Somewhere along the way the company has gotten into nooks and crannies it has no business being in. In fulfilling that need, it may not wind up being valued at $15 billion again, but it certainly will keep the company and users happy in the longer run.