The problem is that smaller doesn't necessarily mean more specific. There is no reason why the gross symptomology used for any of the published case definitions or criteria sets should achieve specificality because none of the gross symptoms used are evidence of specific aetiological processes. The CCC and ICC merely describe random sets within Fukada and Oxford. Reeves has seperate problems and appears to set a random boundary outside of Fukada without obvious justificatio, but in any case it doesn't add to the specificality of the CCC and ICC.
The papers I've been reading by Simon Wessely express a similar view, and for similar reasons. To start with, he views ME/CFS as simply being on the extreme end of a continuum of "fatigue". This is a recurring theme, from his earliest work through his latest.
Then he uses the presumed psychiatric overlap of conditions (especially psychosomatic) to show that differentiating between fatigue and ME/CFS using physical symptoms is meaningless. His reasoning seems to be that since the CDC definition requires patients to have some physical symptoms, it results in selecting many patients who qualify for psychsomatic diagnosis. Of course the CCC/ICC would be even a worse transgressor, in his view.
Finally he argues that the definition of ME/CFS should not include a minimum amount of required physical symptoms to include a patient, and if anything should have a maximum number of physical symptoms, to exclude patients with too many physical symptoms.
It's a rather fascinating perspective, until you look at the results of such a division: patients could retain the same diagnostic label (though he would drop "ME" entirely) if they lack physical symptoms, so CFS would exist entirely of relatively simple fatigue (he's not fond of the chronic requirements either).
But where would patients with physical symptoms go? He does not suggest we be separated into a new category, as some patients or ME advocates might, so the most likely justification for removing us from "his" CFS without a new category would be to put us into an old category - psychosomatic.
Wessely and his colleagues repeatedly express an unquestioning belief that psychiatric illness is present in about three-quarters of ME patients. These beliefs are based on studies using questionnaires focused at assessing psychiatric problems, and the psychiatric questionnaires being used have a rather large number of questions that involve physical symptoms. The result, of course, is that patients with ME or any other systemic physical illness will be seen as strongly psychosomatic based upon giving typical answers for someone with their illness.
This would seem to be a deliberate result that Wessely is looking for. To start with, even the anxiety/depression questionnaire used makes certain presumptions about physical and cognitive capabilities, which could easily result in a borderline score for ME patients (interpreted as positive for psychiatric illness by Wessely & co).
And physical illness can be accounted for even with psychosomatic questionnaires. For starters, the one I took when starting at a "fatigue" clinic included four sections (much like the much briefer General Hospital Questionnaire used by Wessely), but with the caveat that a patient should not be considered psychosomatic when only the somatic section gets a high score - a corresponding high score from one of two of the other categories is also required. This makes more sense (though still doesn't account for depression caused by being chronically ill), because only an idiot would think that a somatic questionnaire is infallible, unless they also think that physical illness does not exist.
Another way to account for potential physical illness in a psychiatric questionnaire would be to compare ME patients to patients with similar symptoms but with an illness that has been unequivocally established as biological. I believe MS patients are used to that end by some researchers. But the Wessely school prefers comparing ME patients to either healthy controls or patients with muscular disease. Healthy controls do not have constant headaches or pain, etc, of course, hence answer psychosomatic questionnaires rather differently than ME patients. And by choosing controls with purely muscular dysfunction, issues regarding sleep and headaches and such (maybe even pain) are rare enough to earn patients a low psychiatric illness score.
While a great many researchers do manage to find mood and psychiatric questionnaires that are appropriate for physical disease, and put them into context by comparing them to diseases with known physical causation, Wessely & Co habitually do the opposite. They don't seem particularly unintelligent, so this may very well be a deliberate choice on their part. At any event, relying on this sort of research to presume physical symptoms are not an important part of ME/CFS is beyond ludicrous.