For those who don't know, I'll mention that rare diseases most doctors don't know about do get classified as functional movement disorders. The problem here is that there is no way within psychiatry to falsify such a diagnosis. Every case I know which has been misdiagnosed had the diagnosis changed after biological evidence of a cause was found. The existence of sane people with mysterious neurological diseases is generally settled at autopsy. If the disease is not progressive a misdiagnosis can persist indefinitely.
Reading the tea leaves to discern organizational intent, assuming any intent exists, is beyond me. I want to concentrate on very legitimate objective questions.
The way this seems to have come together is that the Wichita cohort put together using the "empirical definition" was just sitting there with the label CFS, and somebody decided to use it rather than admit the organization had pissed away money. We then hear that participants will also meet several other criteria with better reputations. Unfortunately, the introduction of new criteria after you have a cohort is likely to mean no more than a check box on a form saying "PEM", and a question by a researcher, "Do you feel worse after exercise?". This approach has obvious defects.
One problem of using a cohort produced by another definition is the propagation of selection effects from the earlier definition, which excluded just about any organic cause of fatigue, even if this was of unknown etiology. Specialists who have been treating "CFS" patients typically find about 3/4 have some kind of thyroid problem, just for example. Even detection of anti-thyroid antibodies doesn't tell us the cause. These patients will probably have been excluded. So will many others with conditions that are labeled rather than understood.
It is also possible to have a range of other autoantibodies affecting liver, heart, muscles, mitochondria, etc. Every autoimmune disease I know about has a problem with diagnostic thresholds, and the typical solution to ambiguity in testing is to choose those levels associated with convenient clinical signs, like the pathognomonic rash in lupus. This will not help in diseases without convenient clinical signs. No law of nature requires such.
Another problem has to do with attribution when a sign is present. Consider Raynaud's phenomenon, in which fingers and toes turn white or blue "in response to cold or emotional stress". The secondary form is definitely associated with connective tissue diseases like lupus, scleroderma or Sjögren’s syndrome. (All generally treated as exclusionary conditions.) There is no good way to distinguish primary from secondary disease. You can read a good bit of medical literature on the subject without ever encountering mention of possible damage to autonomic nerves. One assumption is simply that causation runs "emotional stress" => "reduced blood flow in fingers", not that perception of onset of reductions in blood flow for a condition that can be painful, or even result in tissue damage and gangrene, could produce emotional stress. Please note: autonomic dysfunction which reduces blood flow to extremities may well reduce blood flow to the brain, with perceptible effects long before syncope.
Because of common attribution this kind of sign is often considered evidence of emotional problems rather than physiological disease. Even a patient who exhibits objective measurements showing a substantial drop in pulse pressure during prolonged standing is blamed for deliberately causing something that might only be under voluntary control in certain yogis. It appears that neurally-mediated hypotension is "somebody else's problem", despite objective evidence. This is associated with almost pervasive reports of orthostatic intolerance in CFS. Anyone know how that has been handled in producing cohorts?
These factors make selection of cohorts for "CFS" problematic, and I don't have a comprehensive answer to the problem. What I can say is that the current muddle is nowhere near satisfactory, unless the goal is continued confusion.