There is some precedent for this sort of thing. Being some in ham radio, I’ve worked with and learned about a few different emergency communications systems. And some that are unrelated to ham radio. An interesting project is I2P, but it mostly just piggy backs onto the main internet. Mesh networks are closer to your “our own internet” idea, but they would need to he fairly widespread to be effective. There’s Meshtastic which is an existing mesh network using LoRa tech. Each channel has its own encryption key, and for peer to peer, comms are end-to-end encrypted. There are public channels (which just uses a null encryption key like AA== or something) but if you specify a key, only users subscribed to the same channel with the same key can use that channel. You could technically set up a TCP/IP network over Meshtastic, but it’d be ridiculously slow. But an upside is that there’s already a lot of people and infrastructure using Meshtastic, so you already have a difficult part of building a mesh network done. Unfortunately to guarantee full coverage, you’d have to use an internet gateway to link distant users. There are other mesh networks, but most of the ones I know about use ham radio bands which, legally, need to be unencrypted. (at least in the US) And if you were to try to encrypt comes, nosey hams will be able to track you down and report you. And that’s not a vague threat, a lot of hams are just chomping at the bit to track down violators.

I’ve got some understanding of Tor. I’ve built a pwnagotchi in the past and used good ol’ aircrack-ng, but always got stuck with a hash I can’t crack. From what I’ve seen wpa2 doesn’t go down easily. I have a lot of information, but not much in the way of organizing it usefully, but that’s probably a symptom of having too little or too sparse information.

I just went back to Crimew’s post and started looking at some of those.