If the density and pressure get high enough, there is no longer a stable "neutron" state in which three valence quarks are bound together. You just have a soup of quarks. Calling the quarks "unconfined" is a bit of a misnomer since each of them is restricted to a very small "cell" in space. But they are "unconfined" in the sense of not being bound together by the strong interaction; within their small "cell" they move more or less like free particles.
> the compressed neutrons must be providing some force to prevent them from collapsing further into unconfined quarks.
"Collapse" isn't really a good word to describe this transition. If you add a small amount of mass to the object, it compresses a little more. If it compresses enough, there is something like a phase transition where the quarks stop being bound into neutrons; but the overall size of the object doesn't "collapse", it just gets a little smaller.
> That also raises the question of how the unconfined quarks/gluons provide a force to prevent collapse into black holes.
This is just the Pauli exclusion principle, as has already been said in response to you. It's more or less the same whether the quarks are bound into neutrons or not.
I'd like to point out (for others following along) that the Pauli exclusion principle isn't actually a separate rule (That is, something you'd have to apply after you do 'normal' physics). What is happening with the PEP is that if you start off with a wavefunction that has fermionic symmetry (that is, interchange of two particle swaps the sign of the wavefunction), the evolution via the Schroedinger equation will preserve that (much like it preserves the Integral(|psi|^2)=1 relation). Same for bosons.
So if you're writing a Quantum physics simulator, you don't need to put in a "Pauli Exclusion Rule" step.[1]
[1] Though depending on your representation you may toss one in for numerical stability.