As far as moving the bits around goes, I think that the scuttlebutt protocol handles it nicely. And if you look at the tokenomics of CirclesUBI there's nothing to prevent partition tolerance--you already have the restriction that tokens can only move across links in the web of trust, so you'd just have to add the additional restriction that the parties involved in a transitive-trust-transaction have to be contactable for verification at the time of the transaction. Although CirclesUBI runs on Ethereum (xdai) and would have to be ported to scuttlebutt.
The plan isn't fully fleshed out: you still need to incentivize running nodes and handle cases where bad behavior on the part of node maintainers becomes transparent so that users can revoke trust in them, but my point is that this is not some blind faith in the ability of the community to adapt, but rather something that I have diagrams on my whiteboard for.
Wouldn’t it? This is an unsolved problem and there is nothing that implies it can ‘quickly emerge’.
One of the paradoxes about crypto is that all problems with it can be quickly solved in the future, but somehow continue to exist in the present.