The latter, I suspect. That’s certainly how forming a neutron star works in the first place, because if a star gets so dense that it can form neutronium then the neutronium (which is far more dense than the core was before) can easily keep making more.
It’s a similar story with black holes. Get past the threshold at which it forms, and the process runs away and swallows the whole star.
If a quark soup is more dense than neutronium, then it would be fairly all-or-nothing