Well, I started off by saying "no, food expenditures are not a sunk cost, they are directly related to the amount of work you do". The idea here is that doing x extra work leads to eating x' extra food, where x' is some constant multiple of x (and "food" is measured in energy).
Your response, as far as I can see, says that because Americans already run a caloric surplus, x extra work should only lead to x' - k extra food intake, where the missing k energy is made up by the existing caloric surplus. I responded then and respond again now that this perspective makes no sense; the extra work is an ongoing expenditure and must be exactly balanced by an ongoing intake. For the supposed caloric surplus to be offset against the extra work of extra walking, the person would need to be experiencing constant weight gain beforehand. Nobody[1] is.
You may think that Americans weigh too much, but that's irrelevant; a change in weight is a nearly-pure one-time effect with no influence on dietary requirements. They are eating exactly the amount they already use for the activities they already do.
Am I missing something about your point?
[1] Actually, people with Prader-Willi syndrome will eat if it is possible to eat, no matter how full they are. They will experience constant weight gain if not subject to external control. But they have a serious genetic defect.
