BTW, here is a good paper on ketogenic diets from NIH:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC524027/A couple of interesting things to note:
1) Studies with good methodology (as in have not been refuted in the 20 years since their publication) show that after ketogenic adaptation, peak aerobic power and sub-maximal (60-70%) of VO2 max were fully restored to pre-ketogenic diet levels.
There is no indication that ketogenic diets improve power in these areas. Peter Attia's blog shows marked improvement in these areas, but reading between the lines, the only way that would seem to be possible is if he was training while on his diet. He expresses puzzlement as to why his numbers improved, but it is well known among sports scientists that both VO2 max and Anaerobic Threshold will improve with training. Assuming he was training while on a ketogenic diet, then his numbers really don't show anything useful since there is no corresponding non-keto fitness baseline to compare against.
2) The same studies showed anaerobic performance suffered. In a study of cyclists, their sprinting abilities were constrained by the keto diet. As the paper concludes:
"Therapeutic use of ketogenic diets should not require constraint of most forms of physical labor or recreational activity, with the one caveat that anaerobic (ie, weight lifting or sprint) performance is limited by the low muscle glycogen levels induced by a ketogenic diet, and this would strongly discourage its use under most conditions of competitive athletics."
While the conclusion of "under most conditions of competitive athletics" seems perhaps a bit strong, studies do confirm that anaerobic performance suffers under a ketogenic diet. This should not be controversial.
As I said before, to me the interesting question is whether it is possible to leverage the benefits of a ketogenic diet without incurring the anaerobic penalty. Peter Attia seems to think the answer is yes, so I'll be interested in hearing about his approach and whether it would be suitable for skiing.