While this is interesting, it seems like it would have been more informative with a wider range of macronutrient ratios. 10% carbs for instance or perhaps 50-60% fat. They really tested a pretty moderate range overall. It seems odd that they tested "high protein", but not "low carb".
Also, keeping a hight fat diet down to 8% saturated fat and 150mg cholesterol/1000cal seems unrealistic.
Overall though, it does make a pretty good argument for calories in - calories out reasoning as far as moderate macronutrient ratios are concerned.