Better yet, make a "polymarket" for papers where people can bet on which paper can make it, and rely on "expertise arbitrage" to punish spams.

Doesn't stop the flood, i.e. the unfair asymmetry between the effort to produce vs. effort to review.