From David's post right here: "reasonably convincing evidence that polyphasic sleep may be feasible and practical under some circumstances and/or for some people... I do think that SOME is the operative word here for most if not all of these sorts of things, especially that which is highly dependent upon unconscious or semiconscious phenomena (i.e. altering one's sleep pattern vs. altering one's exercise pattern).
Seems that this might work some (1 to 99%
) of the time for some people and the match of those two factors seems very relevant as well.