If we posit that the observations may have some validity, and set aside the attempt to explain, which is what the `no's addressed, then we could also consider that perhaps it's the degree of hunger that is the trigger here.

Easy experiment-of-one: IF'ers could track the hunger cues they feel before they eat, on a scale of `hmm, did I just hear a tummy-rumble' to `full-out-toddler-tantrum-must-eat-now', then note the amount that they eat, and then whether they feel hot or cold afterwards, perhaps with notes about how much physical activity follows the feed, and track weight as well. A few weeks of this might show whether the theory has any practical application for that individual's life, and they could make a choice then about any modifications they wish to make.

Any statistical relevance? Clearly, no. But our first responsibility is to ourselves, yes? If it makes a difference to the individual, it matters to that person.