The short summary is that aerobic exercise performance in humans is limited by the maximal rate of oxygen uptake or VO2max. Oxidation of fatty acids requires more oxygen per unit of "energy" produced than glycolysis; they are simply less bang for buck (or oxygen). Hence chronically depleted (particularly muscle) glycogen stores are a great recipe for bad exercise performances as you force the body into deriving more energy from fats.

Now, in reality, energy during exercise is always derived from a mix of both fatty acids and glycogen. In general, as exercise intensity increases (you are getting closer to VO2max), a larger ratio of energy will be derived from glycogen. So for your friend, running 54k of mountains is only going to be done at rather low intensity (50-60% of VO2max), where ~60% of energy will be from fatty acids, saving glycogen and preventing hypoglycemia.

Hypoglycemia, incidentally, is also known as "bonking". It's rather unpleasant. On long, high intensity exercise sessions you can't usually replace all the glycogen you deplete through food intake. So if you start with lower glycogen stores through a low carb diet, what you are doing is priming your body for hypoglycemia.

Do you have a scientific source for these claims? Dr. Peter Attia's personal research seems to claim otherwise.

http://eatingacademy.com/how-a-low-carb-diet-affected-my-ath...

Most Olympians are going to be on a high carb diet. What records has Attia set in sports?

Very scientific.

Very crass.

These kind of scary warnings apply to almost no one, basically olympic atheletes in ketosis, and can be treated with a glass of orange juice. For the other 99% of us, ultimate performance is not the goal, and keto is a great way to stabilze blood sugar.

What scary warnings? This is basic exercise science. It applies regardless if your VO2max is 88 or 30, and indeed John Doe is more at risk of hypoglycemia as higher utilization of fatty acid oxidation at lower exercise intensity is one of the primary adaptations gained through training.