Most Linux kernel LPEs --- in fact, the overwhelming majority of them --- don't threaten KVM hosts when exploited in KVM guests.

is there anything good written up on this?

I don't think so? It's not complicated. Most LPEs get you the local kernel. The KVM security model assumes an untrusted local (guest) kernel. To compromise KVM, they either need to be fundamental architectural flaws (rare) or bugs in KVM itself (also rare).