It seems to me that you are missing something critical.
As you say, most people have no established baseline.
ONE possible outcome of this is that if their numbers are normal, they won't receive treatment.
ANOTHER possible outcome of this is EXACTLY THE OPPOSITE--that even if their numbers are normal but they present complaining of all of the symptoms to a high degree, treatment can be justified on the totality of the evidence because, after all, it is unknown what their individual baseline would be--and given the symptomatology with which they present, it could be reasonable to infer that their baseline would, in fact, be outside the norm, and that they therefore merit treatment.
This would be especially true if your last recommendation--that diagnosis needs to be weighted more in favor of symptoms than numbers--were to be accepted.
However, you do raise a point that I find interesting: that maybe the endocrinal and hormonal effects of training are permanent.
I have been arguing that the reason for any deficiency should be critical in determining the permissibility in competition of medication via controlled substances or purified and concentrated forms. This position has so far relied on the assumption that the changes would be reversible by training modification, even if those changes required complete and extended, perhaps permanent, cessation of the training that was their initial cause.
Assuming that the changes can sometimes in fact be irreversible, should that change the above analysis? Put a different way, if an athlete actually makes themselves sick rather than just making themselves overtrained, should they receive different consideration?
That is a tough question to answer. Type II diabetes comes to mind. Bad habits necessitating organ failure, transplant, and subsequent anti-rejection drugs also come to mind.
However, the range of such situations seems to me quite narrow. I would be inclined to answer that if it was demonstrable or likely that the condition was due primarily to training activities, then medicated competition should not be permissible.
Of course, that question is almost impossible to answer. I would hesitate to move the goalposts, to some criterion that would be easier to fulfill, like there being a mere "possibility" that the condition was due primarily to training activities, which would in effect be a presumption that certain activities led to certain conditions. Any such presumption would then shift the burden onto the athlete to rebut that presumption, which does not seem to me like a reasonable burden.
So I would keep the threshold at the above. It would be more of a statement of principle than anything else, as maybe only a tiny number of athletes would ever be successfully banned on any such basis.
Difficulties also arise in the definition of "permanent". Is 6 months long enough to wait? 1 year? 5 years? How could cessation of training activities possibly be enforced?
Maybe adult athletes, upon entering the senior/open ranks, should all have a baseline done at that time, by a governing body or doping authority. If there is an indication at that time, fine. If not, then it will be up to the athlete to prove that future need wasn't occasioned primarily by training activities, whether the future need was "temporary" or "permanent", because it is difficult if not impossible to distinguish between the two.
How would existing users be treated? Not much you could do about that, except establish a baseline upon program implementation. That is the drag of cheating, because some will always get by on technicalities. It is the Mennea effect.