I like this part early on. After deciding fat was to blame, a lot of old studies on respiratory quotient (low respiratory quotient usually means a greater percentage of energy is produced from fat, high more energy from carbohydrate) were used to suggest that the problem in obesity was an inability to adjust energy production to the foods consumed--sometimes they just said obese people had trouble burning fat. Carbohydrate was said to be non-fattening because it encouraged its own oxidation, while increased consumption of fat had only a weak effect. Of course a low carber could look at this and say, wait, carbs might be encouraging their own oxidation, but they're also discouraging fat oxidation.

If you turn it around the right way--however the enzymes, hormones etc. work out, it sort of makes sense for fat to have a "weak effect on satiation" added to carbohydrate. Add carbohydrate to carbohydrate--the fuel available to the body generally increases, since you're overloading the system with excess glucose that must be disposed of. Add fat to carbohydrate--if you're already eating enough carbs to meet most of your fuel needs, then adding fat will only marginally increase fuel availability after a meal. If "getting energy from a meal" has an effect on satiety--and it makes sense from armchair philosophy at the least that it might--fat would have a minimal effect there. But it might have a stronger effect if most calories came from fat--or the system might be more sensitive.

There are studies looking at brain cells that have an effect on appetite. Administered centrally, insulin can decrease appetite, as can leptin. Seems like a bit of a conundrum. An answer I've always liked is that it might just be a difference between systemic and local insulin. Diabetics injecting in one spot for years can develop a lipoma at that spot, calories in, calories out, the spot is exposed to higher concentrations of insulin, so those local fat cells are more well-fed. Same with a bit of insulin directly to the brain--the signal to take in more energy is local, so rather than the threat of hypoglycemia if blood insulin were high, you just get some cells in the hypothalamus that have input to energy intake/use/fat storage being a little better-fed. There are also studies where metabolites/enzymes etc. that are involved in fat synthesis in these cells regulate appetite--which sort of makes sense, usually if fat synthesis is taking place, that's a pretty good signal that energy sources are plentiful.

Maybe I'm getting carried away. I figure anybody with the fortitude to slug through the article might have the attention for my wall of text, too.