Maybe I’m extra brain foggy today, so I can’t make heads or tails of most of your reply.
I think you are focusing too hard on one aspect of the problem.
I have not read the fruit fly paper this time around but my guess is that it describes a mechanism that may provide a regular diurnal clock system through some cyclical use of a metabolic pathway. That would make sense as fruit flies like us are likely to benefit from an 'invariant' 24 reference frame system for timing activities.
But we also know that the human brain superimposes complex varying behavioural structures on this background. I sense the need for a cup of tea at 4.15 pm, not at 11.30 a.m. or 7.30 p.m.. When working my autopilot would manage a vast complexity of timed behaviours, like bringing an outpatient interview to a close after approximately 30 minutes, getting to Grand Rounds, grabbing a sandwich on the way and then rechecking for emails before meeting a PhD student. Complex brain accounting for activity is not a speculation. It is eveyone's universal experience.
Fruit flies may have less complicated schedules but even the spotted crake can be relied on to start calling at precisely 9.15 pm in May on the Poland/Belarus border. But not in April. Migrating crakes integrate over complex mixtures of information to arrive at precisely the same pool year on year.
But you might argue that illness tends to be driven just by simple metabolic levels. I am sceptical. You were not on the forum when some years back I described my wife's illness in some detail because it had some features that overlapped with very severe ME/CFS. It appears to have been triggered by an antimalarial medication, which I understand affects basic metabolic functions selectively, but not entirely so, in plasmodia.
The illness started with hallucinations but these stopped after stopping the drug. What persisted was a psychotic illness that included major changes in sleep cycle and an inability to eat, such that she had to be put on parenteral nutrition to keep her alive. She would always take just enough water to avoid renal failure but was progressively less able to take food and reached under 40kg. Fortunately she was treated and is now entirely healthy.
I do not think for a minute that ME/CFS is the same thing as she had but the DecodeME findings remind me of the sense that the same sorts of systems may be disrupted in both illnesses.
One thing that particularly interested me is that the psychotic state varied during the day. In febrile illness fever tends to have two peaks, with one at night and one in the early afternoon. My wife's illness was worst at night and for a period of weeks at least would almost completely lift around 11.00 a.m. for an hour or so, such that she had insight into her ill state. By mid afternoon a 'different person' had returned. The psychiatrists recognised this phenomenon.
So my analysis is that although we have a CLOCK system that may use a simple metabolic mechanism to standardise the 24 day frame, that frame interacts with extremely complicated accounting processes that assess activities, achievements, and needs way beyond simple physiology (giving a patient full attention or getting to Grand Rounds).
The example of travelling intrigues me because in this case sitting around for hours with nothing to do, but in several different places, with uncertainties about when transitions will occur, appears to shorted my circadian timing frame consistently by about one hour where vigorous physical (off piste skiing) or mental (attending a key small conference) activity does not. So I suspect that even the metabolic clock system can be de-tuned, which it obviously can when you spend a fortnight in a new time zone.
And then of course there is the study in mice suggesting that complement proteins are needed to forget things from day to day, because complement is involved in synaptic plasticity.