Good luck to Branson - I hope he actually gets this off the ground, or at least makes major advances in practical rocket design while he's trying.
But the last few projects like this - Teledesic, Iridium, a couple of other important ones I forget - all ran into problems with markets, with costs, with technology, and with government regulation (both censorship and spectrum-control.) One of the cool things about satellite phones and data was that you could access them from anywhere in the world, even places without much infrastructure, but the problem was that they cost a lot more than terrestrial infrastructure in densely populated areas (so you couldn't make much money where there were lots of people), and sparsely populated areas are mostly poor farmers (so you couldn't make much money there), so what you really had was a niche market that cost you billions in upfront infrastructure. It's also hard to get high bandwidth from solutions like this (though lots of applications don't need to be that fast.)
Governments were also a problem, because many of them didn't want unregulated speech, not subject to wiretap, competing with monopoly or ex-monopoly local telecom providers. Remember when Blackberry was only allowed to sell their phones in India if they provided a nexus for wiretapping?
There have also been half a dozen announcements over the last decade or two about balloon-based projects, with blimps or weather balloons or tethered balloons or whatever providing low-altitude radio towers, which can deliver a lot more bandwidth (because they're close and can carry a lot more power), but somehow none of them ever turn into reality. (Good luck to Google and Facebook on those.)