re: corruption, the impromptu government in my scenario only exists because a group of people decided to form it by collectively deciding the same way re: which money they'll accept and which they won't. It is legitimate so long as they continue to agree. If it stops being legitimate, then it stops being effective at the problem it was created to solve, people stop agreeing, and
poof, it disappears.
The goal is only to make this kind of resistance possible. If it's the wrong tool for the job, stop using it and use something else.
re: incumbents, the assumption is that they're so sclerotic that they can't respond to novel scenarios. If this weren't the case, you could've just used them to deal with the driniking water problem in the first place.
re: fungibility and barter, the idea that "before we had money, we had barter" is often repeated in economics classes, but the anthropological evidence doesn't support it. Before the Roman empire violently asserted that caesar-face coin was more valuable than the metal it was made of, we did more or less what I'm describing. People with shared interests kept track of debts between each other, and people without shared interests refused to do so. It didn't scale super well, but now that we have computers, it can.
re: fungibility and kyc, there are plenty of people who are in jail for at one point having had the wrong kind of money, because it's not fungible.
re: a meshed graph containing everyone in the world
Well yes, that's how we treat money today, but I don't think it's sustainable. It concentrates too fast into the hands of too few. Why play a game in which you were born into a losing position? My generation decided to play, but I don't think that'll keep happening, because every generation has worse and worse reasons to. The French got fed up with their royalty, the children of the future will get fed up with their Quadrillionaires. The French chose violence, I'm proposing something else. A way to make old money into not money without taking anybody's head.
If you opt in to something like this, all that remains is new money--money issued with a purpose (we have this already, money enters the system when loans are granted, those loans were granted because somebody at a bank thought something was worth doing, a mining project perhaps). You don't have to track every single transaction, just the ones where money is created. I think people should have the ability to look at money and decide if its original purpose was a worthwhile one (nowadays the loan reasons are hidden from us). This would create better behavior among those so audacious as to create things and propose that they be treated like money.
For inflation reasons, money will also need to exit the system, so it's not going to hang around forever. As for whether you need metadata on every hand that is passes though from creation to destruction, or whether it's good enough to accept/reject anonymously handled money based solely on the conditions of issuance... that's up to the people. If they deem an abstraction legitimate by being willing to accept it as money, so be it. If not, well I guess we'll need bigger databases or more aggressive demurrage schemes such that the histories don't become unwieldy.
I realize that this all seems a bit outlandish, I must seem like a crazy person to you. But it's clear to me that our system has some very damaging design flaws, and I can't tolerate watching the subsequent harm without at least spinning my wheels now and again in search of an alternative.