I think Dr Ludwig made it clear that the “low carb beats low fat” claim only applies when people actually follow the diets. This is best shown experimentally in feeding studies and short-term trials. Once they go back to eating as they used to the difference is attenuated, but I don’t see how that refutes the claim.
Guyenet’s point #10, about lean high-carb populations, needs I think a better response than saying that activity explains it (which is just CICO reheated).
The Kitavans, for example, stay lean on high starch diets and have low insulin levels — this does not refute the hypothesis, but perhaps reflects another mechanism such as the acellular carbohydrate hypothesis, or the beneficial effect of a diet that is also limited in essential fatty acid PUFAs on the first phase insulin response.
Populations that are lean on white rice may reflect the fact that storing fat, like any activity, requires adequate micronutrition and protein, and diets that are relatively deficient in B vitamins such as thiamine may not favour fat storage. For example, alcohol is rich in calories but alcoholism, apart from beer drinking, does not usually lead to obesity if the non-alcohol diet is inadequate.
True low fat diets do not predispose to obesity because even if insulin is high it cannot store much fat unless there is preformed fat in the diet. De novo lipogenesis contributes to fat storage on normal diets but cannot supply enough fat to compensate for a near-totally fat deficient diet. Many cells require fat as a fuel, all require it for structural and signalling purposes, so DNL fats will be used up as fast as they can be synthesised on such a diet.
None of this contradicts the carbohydrate-insulin hypothesis of obesity — it may be conditional on many contingencies, but these conditions are pretty reliably fulfilled in our modern world.