The internet is successful because the internet protocol is a powerful medium. Service providers can imagine many ways of splitting up the medium to make more money, however the corporate conception is necessarily short sighted.
On the long term, corporations evolve in fundamental ways. They can open, merge, split, be acquired, and close. This is appropriate to individual corporations including capital market service providers, but a headache for the consumers of infrastructure providers.
For profit internet service seems to work well enough on the short term. However, in the past decade we have seen that most internet service consumers have no access to the multicast protocols. And of course there's the famous case of Comcast versus BitTorrent.
Perhaps corporations are not capable of being responsible for basic infrastructure over the long term, due to the need for continuous growth in the economic rent paid on public equity capital.
In other words, if for profit internet service providers can't do the business they entered into then other forms of internet service infrastructure will be needed. The public postal service is not the only alternative.
Wireless mesh networks have been developed by municipal governments, and individuals who install a mesh node somewhere.
If every house had a mesh node, then internet would route normally between houses and collective entities like cities and municipalities could connect the mesh to the conventional internet with a large, shared pipe.
This has been done in a number of places including California and Massachusetts.