It ignores metabolism, there is evidence that the presence (or lack of) certain enzymes in the digestive system can drastically affect ability to absorb calories, etc.
It is true that there are many factors that affect how a calorie is digested and stored, it is a complex system.
But it is also true that there is no other source of energy for the body than mass. There is no metabolic process that can produce energy for more than a few seconds without a reaction that consumes mass (the carbon in said mass being breathed out as CO2).
Just for fun, I'll paste my reply to another comment here:
Do you genuinely believe that metabolism doesn't vary across individuals, or for an individual over time? Do you think the bodies of all individuals have precisely the same ability to both absorb calories from ingested food, as well as produce motion with the exact same efficiency? Do you believe enzymes present in the digestive system have absolutely no effect whatsoever on the calories in / calories out calculations? Do you think every calorie you put into your body is converted into energy?
I'll add an extra one for you: in controlled lab tests, would all subjects lose weight precisely as predicted by a formula that considered their diet and physical activity?
This page has links to 20 controlled studies, and the calories in - calories out hold. It is when people are allowed to self report, the problems begin.
Sure, everyone's caloric requirements are different, but everyone will lose weight if in caloric deficit. It is just the question of figuring out what will be deficit for this particular person.
Do you need a subscription to read the full articles? For one thing, I'd like to read how they are measuring "energy in", are they subtracting calories in the subject's excrement?
Some articles are open access. In one case, for example, they used bomb calorimetry:
Bomb calorimetry performed on eight batches ofsimilar formula prepared in the research kitchen of the Rockefeller University Hospital between April 1988 and July 1990 had a coefficient of variation for Id/g of 1 .9%.
Again, I've seen little evidence of excreting unprocessed calories being a common case. And if we're looking at weight loss anyway, excreting some calories would even be an advantage.
