Where have those physics been documented?

I want to see the same person:

    'immobile
    'sitting most of the time, with short commutes and shopping trips
    'intensely exercising 1 hr a day
    'actively working out throughout the day

while maintaining their current diet and weight.

These aren't independent variables. The point is that a half an hour run only burns around 300-400 calories, a deficit which could more easily be accomplished by diet. Also, it's possible that the run would increase your appetite.

The point isn't that exercise _can't_ work, it's that aiming for dietary changes has a higher average ROI than exercise.

"these"? I don't know what that refers to.

No one has argued that there is a lower than average ROI (you get paid for not eating?) from exercise.

It's not physics, it's biochemistry.

But it's more complicated than that. Genetics cause one's predisposition as to how responsive a person is to cardio; see the University of Bath study and/or Michael Mosley's "The Truth About Exercise".