f@h studies protein folding which relates to a whole bunch of different diseases...including many forms of cancer. It's up to the Stanford folks to interpret those results from the program.
The reason I'm down to one machine is because something weird happens (maybe between the SMP client and debian) where the time (as reported by date) stops and possibly as a result, or maybe as an additional symptom of an underlying problem, folding doesn't go anywhere. Not really sure on the details but it happened on my previous debian SMP machine and that had totally different hardware other than graphics cards.