People move from platforms all the time. Before FB there was MySpace, before Google there was AltaVista, before GApps you'd probably use MS Exchange etc etc. There are many examples.
What's needed is software infrastructure to make it easy to build decentralised systems -- everyone should have their own piece of the cloud. On top of that is the need for a compelling use-case (beyond privacy) that will spur word-of-mouth and user adoption. All this has happened before, and all this will happen again.
The sad thing is all of the technology essentially already exists in some form or another - it's just that no one put it all together in an easy-to-configure box with a web interface. RSS for updates, chat over XMPP, messaging with SMTP+TLS+PGP. At this point it seems like most of the tools for the backend are done and the rest is would just involving plugging them together, putting a nice interface on top and acquiring the user base (I know - understatement of the year).
There is movement in this area - see sandstorm, owncloud, disapora, etc. Slap it all in an Odroid and you have control of your data again. But people won't pay for it.
I tend to agree, but really nobody will respect your privacy but yourself, which is why this should be on your own hardware (or at least on a VPS). A small box under my couch hosts my site, cloud storage (owncloud), and if I could get away from it the RSS feed that basically is facebook, and it's quite nice. It was too hard to set up, though, which is why sandstorm looks great.
The people most likely to be early adopters of such a platform would distrust a solution hosted by a 3rd party. Why? Because if you don't control the system, it can still be used to spy on you. Even a decentralized system with Wordpress-like companies that are federated are not immune.
I hear this a lot, that people move platforms "all the time", but it's not really true as it relates to Facebook. Before Facebook, the vast majority of users on social media (MySpace), were teenagers and young adults. Social media only exploded to your parents and grand parents and everyone else with Facebook. It's much harder getting those folks to switch, and the network effect is that much stronger with the ~30x larger user base. Switching off something like Alta Vista to Google isn't even comparable - there are basically no switching costs in that case.
I'm not saying it's impossible, but the stickiness and network effects of Facebook are pretty much unprecedented.
You need to factor in time. As an example Microsoft isn't treated the same way it was years ago and IBM isn't the same as before Microsoft.
I'm not suggesting FB will become obsolete but people will use other services if they see value elsewhere. If that weren't the case then WhatsApp/Instagram wouldn't have grown so fast. Network-effects does not mean something is immune from disruption.
What's needed is software infrastructure to make it easy to build decentralised systems -- everyone should have their own piece of the cloud. On top of that is the need for a compelling use-case (beyond privacy) that will spur word-of-mouth and user adoption. All this has happened before, and all this will happen again.