Here I'd add "in the liver." Otherwise you might end up in the "fat burns in the flame of carbohydrate" territory--the liver's inefficiency that produces ketones is special to the liver. Gluconeogenesis is an oxaloacetate sink. You don't have to be in ketosis to be burning lots of fat--but ketosis shouldn't be mistaken as showing a more general inability to efficiently burn fat.

I've seen people on facebook go so far as to suggest that ketones in the urine show that a person isn't fat adapted, something that's really not true.

I actually don't think post-meal--at least the few hours after a meal--insulin is a real big issue for bodyweight. Fasting insulin might be. But define fasting. If your liver glycogen is depleted, your 'fasting' insulin is likely to be lower. That's postmeal, but it's postmeal after return to fasting glucose levels. Fasting glucose only goes so far in showing your metabolic state.

Dietary fat mixed with carbs keeps insulin elevated longer, even if the peak is lower. For a given level of carb intake. Fat alone won't raise insulin much at all.

Earlier high carb low fat researchers talked about a glycogen-stat--people eating to maintain glycogen at a certain level. Fat didn't add to this so sort of came along for the ride. So, don't eat fat. The other side is--people eat to maintain the glycogen they're accustomed to. Fat still comes along for the ride. But you can take the wheels of by becoming sort of unaccustomed.

For me--the reason we have the insulin and carbohydrate and palatability hypotheses is to explain observed events. We can pick at the hypotheses--but still need to explain the observed events--such as, overweight people often spontaneously losing weight just eating low-carb foods (which may be less insulinogenic, but that doesn't necessarily make that the mechanism) to appetite. You can never use hypothesis or even more established theory to disregard observation.

Also for me, personally--I ate Atkins with unlimited protein for years, that got me down around 170, I had to struggle to get down to 160, tried a 4:1 very high fat ketogenic diet after painstakingly dieting down to the mid-150s--and found easier maintenance there. That's 90 percent fat by calories--now I'm more around 80-85, but still haven't climbed out of the 150s, usually between 153-156, and I've been there for years.

Glycemic index nonsense had full-sugar ice cream looking better than plain baked potatoes. I'd take the potato, fortunately I don't have to take either. But even there--okay. Bigger glycemic response, maybe higher initial insulin. You return to baseline--free fatty acids may end up higher than before eating. They get suppressed by the insulin, mechanisms to dispose of glucose are upregulated to handle all that glucose washing into the system--but then as the incoming glucose dwindles, there's a counterregulatory hormonal response to prevent a hypoglycemic event.

I do think it's still much about the insulin--it's just that it's not as simple as we'd hope for.