Something I've seen Gary Taubes talk about--I think after one of his lectures after he wrote How We Get Fat-- is the danger that, if the studies went the right way, and the Carbohydrate hypothesis became as entrenched as the low-fat, calories in/calories out thingum did in the past--instead of looking at obese people and saying--well, look what they did to themselves--gluttony and sloth--people would assume they were gorging themselves on carbs. That we'd just start blaming anybody whose obesity didn't simply resolve with a low carbohydrate diet, still blame the victim, just with a slightly different accusation. I think it's fair to think that a low carb diet might help somebody--just as if you knew somebody with epilepsy or cancer, it would be fair to think that a ketogenic diet might help them. Just like with the epilepsy and cancer, you really can't know for certain whether and to what extent a person's issues will resolve with the diet, though.
Atkin's fat fast for the extremely metabolically resistant--the plan went, one week of 1000 calories, ninety percent fat, then a week of induction, then back to the fat fast... he didn't think the science was there to safely advise people to do longer term fat fasts. But even the induction diet... done right, by my reading, it's no different from nutritional ketosis. He doesn't spell out protein/fat ratios, but he has people measuring ketones (by the only method available to him at the time). And in DANDR, he reminds readers that two thirds of protein can be made into glucose and only ten percent of fat (the glycerol fraction). I have little doubt he would have been right on that blood ketone meter thing, if he'd had the chance. Induction was designed to replace the two day fast that earlier low-carb diets had been initiated with (I think Donaldson's was?) and Atkins called it a diet that was "effectively" zero-carb.
So essentially-if my reading of Atkins is correct--for metabolically resistant people, what we now call Nutritional Ketosis (tm), interspersed with 1000 calorie Nutritional Ketosis Extreme. I'm not sure how you can read that recommendation and go on blaming the victim, even if you're a card-carrying member of the ketogenic cult, as I am.
Even if we ignore everything but insulin and carbohydrate. If you accept that, by ruining ketosis, very small amounts of carbohydrate can prevent weight loss on a low carbohydrate diet--and some of the therapeutic benefits of the diet that depend on ketosis itself--well, this pretty much predicts that some people might not get there, however draconian the dietary measures taken. Because the less dietary carbohydrate or protein it takes to prevent ketosis, the smaller the bump in glucose production from endogenous sources it should take to preserve glucose metabolism/prevent ketosis.
I think the only thing really wrong with the Insulin hypothesis is that people want to stop there. I'm also not crazy about calling it the Master Hormone. Yes, remove insulin, and fat can't sustain itself. With an insulin-centric view of things, people will say, leptin deficiency makes people/animals fat because when leptin/leptin signalling is deficient, insulin goes up. Which is fine. But who's the master here? The first cause described here is the low leptin, not the insulin. Other things that could happen--high glucocorticoids, and other counter hormones that raise glucose--forcing the pancreas to increase insulin production to keep blood glucose from getting too high. Maybe the insulin causes the obesity--but what made the insulin go up, here? Carbohydrate drives insulin drives obesity would be more correctly stated "Carbohydrates and other stuff that drives insulin drives obesity."