Luther:

1. No matter how many white swans you cherry-pick, I've already posted over a dozen black swans -- so I'm not sure what you're trying to prove.

2. None of the studies you've linked contradict any of the studies I've described in this series.  They only prove that variations in macronutrient balance, away from the extremes, don't change much.  (For instance, the lowest carbohydrate content of any diet was 25%.)

3. The sentences you cherry-picked often don't mean what you imply they mean.

However, I thank you for bringing these studies to my attention, because they'll be useful in a future installment about metabolic individuality!

Let's go through them one by one.

The first study you posted (McDevitt 2000) fed subjects a normal, high-carb diet -- and only then overfed the subjects with fat or carbohydrate.  Result: the macro percentages didn't change all that much.

Quote: "The control diets provided 48% of energy as carbohydrate, 40% as fat, and 12% as protein. The [carb-overfeeding] diets provided 50% of energy as carbohydrate, 42% as fat, and 8% as protein. The [fat-overfeeding] diet provided 32% of energy as carbohydrate, 60% as fat, and 8% as protein."

Furthermore, it didn't even measure weight gain or loss!  All they did was force the subjects into a room for five days, force them to exercise identically, and measure both heat production and RER to approximate fuel usage.

Anticlimactic result, which doesn't contradict any of the studies I've cited: changing your diet form 50% carb/42% fat to 32% carb/60% fat doesn't significantly change how much energy you expend over five days.  Raise your hand if you're surprised.

Next, Lammert 2000.

"In phase 3, the C-group received 78 % of energy as carbohydrates, 11 % of energy as protein and 11 % of energy as fat, while the F-group received 58 % of energy as fat, 11 % of energy as protein and 31 % of energy as carbohydrate."  So we're talking about the difference between a high-carb diet and a moderate-carb diet.

Anticlimactic result, which doesn't contradict any of the studies I've cited: changing your diet from 78% carb/11% fat to 31% carb/58% fat doesn't significantly change average weight gain.  Again, raise your hand if you're surprised.

Next, Golay 1996.  I don't have access to fulltext, but the abstract tells us enough.

"The patients were assigned to one of two groups that received either a low (25% CHO, n = 31) or a high (45% CHO, n = 37) carbohydrate hypocaloric diet (5.0 MJ/d, 1200 Kcal/d)." 

First, I note that the 25% carb group lost 19% more weight than the 45% carb group, though that wasn't deemed "significant".

Anticlimactic result, which doesn't contradict any of the studies I've cited: changing your diet from 45% to 25% carb produces a 19% improvement in weight loss, which the authors don't believe to be statistically significant.  

Finally, Davy 2001.

Again, the diets weren't all that different: "The specific macronutrient composition (protein, fat, carbohydrate) of each diet, as a percentage of total energy intake, was 15/30/55 for the mixed diet condition (M), 15/60/25 for the high-fat diet condition (HF), and 15/15/70 for the high-carbohydrate condition (HC)."

And again, this study didn't even measure weight gain or loss: it measured RER using indirect calorimetry.

Anticlimactic result, which doesn't contradict any of the studies I've cited: changing your diet from 70% carb/15% fat to 25% carb/60% fat doesn't significantly change how much energy you expend over 24 hours.  Again, raise your hand if you're surprised.

Finally, I've already addressed Buchholz 2004 in the comments to Part I, but I'll expand on them here.

Let me highlight these sentences, because they're so monumentally self-contradictory – and highlight the deficiencies of the calorie model so well.  Quote: "Diets high in protein and/or low in carbohydrate produced an ≈2.5-kg greater weight loss after 12 wk of treatment. Neither macronutrient-specific differences in the availability of dietary energy nor changes in energy expenditure could explain these differences in weight loss."

In other words, the CICO model, while trivially true in the physics sense, is completely inadequate to explain observed reality.  (By which I mean “Attempting to adjust the “calorie”-based calculations for all the factors I discussed in my article is apparently, in practice, impossible.”)

“As such, a calorie is a calorie.”

Wait, what?  They've just admitted that it isn't.

“Further research is needed to identify the mechanisms that result in greater weight loss with one diet than with another.”

And now they admit again that the calorie-based model fails to explain observed reality.

Conclusions:

1. Your references show that varying carb content between 75% and 31%, while keeping protein content constant (between 8% and 15%), doesn't have a significant effect. 

2. These references do not contradict any of the citations in this series.

Result: My points stand.

Catherine:

Read my reply above.  Luther isn't making the point he believes he's making.

JS

eddie:

As I said way back in 2011, in Eat Like A Predator, "Abstracts and conclusions often misrepresent the data."  It's reasonably easy to design an experiment to get the results you want...and if you don't, it's easy to "adjust" the data and move the goalposts until you do.

One future installment of this series will cover the topic of "When is a calorie a calorie?" i.e. "Under what circumstances do dietary changes not matter?"  One of the main ones seems to be "Swapping carbohydrate for fats that aren't coconut oil (i.e. high in MCTs) doesn't seem to change things unless you're at the extremes of carb or fat consumption."

JS