Fung doesn't make a really coherent argument about what goes wrong with Atkins over the long term. He argues that long-term failure of adherence to Atkins is a function of adaptation to the diet. Over and over again he makes arguments for intermittent, or pulsatile solutions -- ultimately fasting. In particular he cites this study as an argument that intermittent fasting is more effective than continuous calorie restriction. By proxy he seems to argue that this is what is missing with the low carb approach - IE because low carb is meant to be a long-term, continuous behavior, it fails due to some kind of adaptation or development of resistance.
He may be right about the fasting, but I don't find the argument that "Atkins fails because it is too restrictive... what you need to do is add fasting" to be particularly compelling. Peter Attia in an interview said that the reason he fasts is to give himself a bit of flexibility in his diet over strict "keto". This is a much more cogent argument.
"Well I mean if we are going to get really technical, the benefits are overstated, or only proved in animal models... So I don't know if those benefits would extend... So for me what it comes down to is that it's just an easier way, it gives me much more liberty with what I eat during my eating window, I don't have to be nearly as restrictive when I'm eating if I have that period off..."-- Peter Attia MD
