I love her work, QubesOS is great. Yet it never ceases to amaze me how few citations there are to other academic work in this domain :) "Provably" secure operating systems seem to go back to at least the 1960s in the literature.
Back on topic, I guess this is why more exotic secure operating systems (seL4/Genode, etc) are focused mainly on "smaller", non-x86 CPU architectures (ARM/PPC/etc).
