To cut through the not particularly helpful analogies, the core result (which the article finally mentioned near the end) seems to be finding a situation where

To cut through the not particularly helpful analogies, the core result (which the article finally mentioned near the end) seems to be finding a situation where the Euler equations work, but only for a finite time before hitting a singularity. Previously, the only known cases either hit a singularity immediately or never hit one, i.e. every case either works forever or doesn't work at all.