So FWIW I just told john two things (we work in the unix room here at sandia labs ... we even have a pjw poster). I just get in later because he is the industrious intern and I am the grizzled old guy.
First: the boot process is dying. One easy thing to do is set up so that when you boot it drops you into an rc running out of ramdisk, and just poke around, see if you can do things to the disk, run paqfs (can even do it under acid, probably) and so on. You can learn a lot. Second, the usual, debug prints in paqfs ... ron