The majority of people with ME/CFS do worse with alcohol after developing the disease. (See this
poll, for example.) And I previously mentioned
here one possibility for this which is related to S1P and phospholipase D (PLD).
Specifically with respect to acute alcohol ingestion, this is likely to be even worse for most people with ME/CFS and I personally wouldn't distinguish this. As far as your implication as to how a small subset of ME/CFS patients seem to benefit from this acute ingestion, I don't know.
With respect to FXR, since alcohol antagonizes this enzyme, and since inhibition of FXR leads to decreased synthesis of ceramides - from which S1P is made - then that seems consistent under a simplistic version of my model for alcohol making ME/CFS symptoms worse:
Inhibited FXR-->Low ceramides-->Low S1P/Impaired S1P signaling-->Under-activated Akt/mTORC1 pathway-->Impaired PDH complex (Fluge & Mella)
(It's getting too far off topic, but one research group in south Florida seems to be focused on how S1P can be useful with respect to protection or restoration of blood brain barrier (BBB) damage caused by alcohol, so the corollary for people with ME/CFS would be that not having enough S1P would make us more vulnerable to that damage. (E.g.,
reference.) So this might be an additional/alternative mechanism for the bad effects of alcohol in addition to what I mentioned about the PLD enzyme.)