As someone with a BMI that typically exceeds obese by a percentage point or two, I am wary of aggregate diet data.

I'm fat not because of snacking, but because of 1500, 2000, and 2500 calorie meals.

N of 1 here. Take it for what it's worth, I'm skeptical of averaged data like this. There is simply no way to tell whether it's outliers like me skewing the data, or whether it represents reality.