You're talking about Gridcoin. It didn't do too great economically, for reasons that become obvious when you think about it for a while. It was a great way to fund research, but you can't actually build a global economy on BOINC tasks.
More of a decentralized compute marketplace rather than a PoW/emission mechanism, but if general compute is what you're looking for, then there it is.
PoW requires that the "work" have a few special properties, which makes it really difficult for it to be something useful. I feel like we're kind of past the whole "looking for a better PoW" phase though now that most consensus runs off really mature PoS mechanics.
Try thinking of a useful problem that allows everyone in the world to endlessly generate puzzles worth solving, whose difficulty can also be ratcheted up and down to regulate block timing.