Here's some food for thought in regards to the set-point theory and calorie counting. It is my contention that the set-point theory is an attempt to explain why calorie counting has such a horrible track record as a weight loss method.
Assume an average person consuming 2,000 kcals per day. That's about 750,000 kcals per year. Now, current calorie-counting theory states that fat cells grow or shrink at a rate of 3,500 calories per pound. That is to say, a calorie surplus of 3,500 calories, in excess of basic metabolic needs, will result in the growth of a pound of bodyfat. The same goes for a calorie deficit.
Are you with me? A pound of bodyfat = 3,500 calories. In theory. Our average person is consuming 750,000 calories per year.
Question: When calorie counting, how accurate would this person need to be in order to maintain their current weight? Plus or minus 5 lbs. per year?
Consider that the vast majority of people do so without even counting calories.
Answer: They would need to accurate to within 48 calories per day. That's 2% accuracy. Outside of a laboratory that, ladies and gentlemen, is impossible. Misjudge that ice cream by 1 tablespoon and change the amount of walking done by 10 minutes and already it's off by that much. If Burger Boy puts an extra half-tablespoon of mayo on the sandwich, calorie counting for the day is screwed. And what if that chicken is 6 ounces instead of 5? Oops, thats 50 calories difference.
Does anyone think that a triathlete like Eugene Allen can calculate his calories expended to within a 2% degree of accuracy? Over the course of the 3,000 or 4,000 or 5,000 calories he may burn in a day? Eugene, how much has your weight fluctuated from year-to-year?
Is there a self-regulating system within the body that - under normal circumstances - keeps bodyweight relatively stable? The math suggest there is. Again, consider the assumption that MOST people don't gain or lose weight at a high rate, usually 5 lbs. a year or so over the course of their lifetime. So, if calorie-counting is so critical to achieving ideal bodyweight, then how is it that so many people who DON'T count calories manage stay within 20-40 lbs. of their ideal weight for decades on end?
No one, maybe not even most labs, can calculate BMR and daily calorie expenditure and daily calorie intake to within a 2% accuracy. Yet that is exactly what the calorie-counting nazis like the American Council On Fitness and Nutrition are asking people to do.