First of all, I'd like to point out that very low carb (VLC) and ketogenic diets can be effective therapeutic tools for treating many neurological disorders. I touched on this briefly a while back in my podcast with Emily Deans, and initial studies on low-carb diets and mental health have shown promise. (1, 2, 3, 4)
Because Dr. Perlmutter is a neurologist, it makes sense that he would be a proponent of low-carb diets for his patients based on these therapeutic effects.
However, recommending a low-carb diet as an intervention for sick people is very different from promoting it as a preventative measure for the entire population, which is what Dr. Perlmutter does in Grain Brain. His approach would be somewhat akin to recommending that everyone go on the Autoimmune Protocol to prevent autoimmune disease, which would be unnecessarily restrictive and unhelpful.
Comment: While Chris Kresser is a well-respected researcher in the Paleo world, he is showing his bias here. There is a great deal of evidence that a ketogenic diet is beneficial for many conditions, not just neurological and that yes, it is indeed beneficial in a preventative manner. For a comprehensive review of the literature, check out "The Ketogenic Diet: An Overview".
It's important to realize that just because a low-carb diet can help treat neurological disorders, doesn't mean the carbs caused the disorder in the first place. While I don't argue with the idea that refined and processed carbs like flour and sugar contribute to modern disease, there's no evidence to suggest that unrefined, whole-food carbohydrates do. In fact, there are three compelling reasons why this is not the case.
Comment: It's a misnomer to label a ketogenic a "canned approach". There is a great deal of personalization that can be accomplished on a ketogenic diet, as will naturally happen with any diet. In fact, Kresser's Personal Paleo code could likely be adjusted to allow for a ketogenic approach.