Including light (ie - free space optical networks, which are usually commercial, but the Ronja project shows they can be DIY'd too)?
I've often thought that mesh networks are where things need to move, but they suffer from a larger problem here in the States:
Chokepoints
Essentially, there are large gaps between cities where there isn't enough population, or there are geographic features in the way (or both) that preclude mesh network relays from being setup by anyone other than a corporation which has the resources to purchase the land needed, and either install fiber, microwave relay towers, or whatnot - and maintain all of it over time.
Sure - you could probably mesh network most of the NE of the country. Maybe even California (though there'd be bottlenecks). But how do you jump from LA to Phoenix, or from Phoenix to Las Vegas or DFW? How do you link any of that to the east coast? How do you get this mesh network to the interior of the country?
Not likely to happen without a lot of money and effort.
What I can see happening, though, is a return to lower-bandwidth communication networks - ie, dust off that old BBS software! Dial-up, or low-bitrate comms (which can pass easier thru congested or low-node count mesh networks) would likely become the way to do things.
But I doubt it would become mainstream, as most people just don't seem to be able to live without their streaming TV channels and bookface feeds. But maybe that would be ideal; the hardcore and tech conscious using these lower-level channels to figure out a way to fix the problems in the mesh, so that one day all could share it.
Worst case, the signal to noise ratio would go down, and we geeks could finally have a space with intelligent discourse and fewer trolls.