I once started writing a distributed social network, and then life took me on a journey. (I'm still finding the time, though.)
The reason why we don't have one yet is that writing a distributed social network is HARD. It's a much harder problem than inventing the web or email, because the security stakes are much higher. The consequences of spamming and spoofing are even worse than what we see in email; thus an author of a distributed social network needs to solve this problem early in the process.
Another problem is encrypted communication. Https requires buying certificates, thus a well-designed distributed social network needs a means of key distribution that allows a casual server operator to get running without purchasing a certificate.
When I met the Diaspora team, they were very ambitious; but they just weren't experienced enough for the task. Something like a distributed social network requires a team with significant experience, much more then a group of fresh grads will have.