That does pose some interesting puzzles.

Perhaps the fact of moving changes how the body partitions nutrients. Or handles hunger. Or some other factor. Or the specific modalities of exercise (brief, high-intensity vs. long, low-intensity) matter.

The study appears to rule out one possible mechanism (total calories consumed). There's still the fact of bodyweight / body composition differences.

If your test fails to account for observed differences in phenomena, then your test failed to identify the cause(s).

Just as the sick man who goes to the doctor for a battery of diagnostics and is told "we can't find anything wrong". The problem isn't in the state of the man's health, it's in the state of the doctor's diagnostics.

"... they expend on average no more calories in a day than the adult population of the industrialized world." Did they take weight difference into account? is it per kg or per person? I find the article lacking in a few important details...