In lower firmwares a user could hijack a process and the PS3 would continue to spawn new processes, but in newer firmwares if a process has a lock the hypervisor will prevent any new processes from starting until that lock has been released.
(I'm guessing a similar thing to SQL when it has a lock - however SQL will choose a deadlock victim whereas hypervisor will just sit and wait it out. To prevent any code being run)
The Forrox Author should have no credibility whatsoever. His stuff is the grandfather of all PS3 fakes. Nothing was adjusted to stop Ferrox working, because he never had anything working in the first place.
What nonsense crap he speaks now. HV privileges were adjusted in 1.6, what a bunch of crap. The HV runs at HV privilege. End of story.
Note: I am not in any way supporting the notion that the other stuff here being referred to is real.