My own conclusion: exercise and diet are overrated, I'll still do sleep experiments because I don't want to spend 1/3 of my life unconscious in bed. Things like uberman, everyman, siestas, melatonin supplements, and modifinil are all intriguing.

The writing of this entry bugs me. Generalizing from sample size of 1, "confidential sources" (he won't even link to papers apparently already published, all years ago!), and the sentence "For lack of space, I cannot explain here the ideal sleep architecture."

Taking the last stat to be true (Sleeping for <=5 hours per night leads to 39% increase in heart disease.) that's still only a change from something like 0.6% to 0.8%. (Initial percent from memory, so it may be off.)