I mean, as long as we're fabricating fantastic situations we can maybe come up with one where PoW actually makes sense. So the thing about PoW is that it doesn

I mean, as long as we're fabricating fantastic situations we can maybe come up with one where PoW actually makes sense. So the thing about PoW is that it doesn't matter how much is actually spent on the "work" part as long as you can't feasibly do more. So let's add onto our climate change apocalypse some light nuclear war -- not enough that it dominates the end of the world as we know it, but enough that it destroys the infrastructure needed to produce the most specialized mining hardware, and maybe some of the larger datacenters. Now you have plenty of scavenged ordinary hardware to keep the internet running, plentiful power (since in this hypothetical the population has shrunk dramatically without much infrastructure loss) and a finite supply of mining hardware. This last would prevent the runaway problem of PoW -- you could string thousands of regular PCs together to compete with the good hardware, but the return would be low enough that it wouldn't be worth while to try to expand the mining pool until the old hardware started to fail completely or the new infrastructure was built to fab new hardware -- you might have a few decades before the old incentives to grow the pool kicked back in. Granted, this is all pretty far-fetched, not the least the social component of people wanting to use cryptocurrency as a widespread unit of exchange, but hey, I'm work-shopping a story set in a dystopian future here, not trying to sell Bitcoin.