How does the person know that the improvement isn't a fluctuation in illness caused by something else?
I got as far as buying some naltrexone with the intention of making up some solutions, but by the time I had read around the evidence, I lost interest. I haven't seen anything sufficiently convincing. The balance of anecdotal reports seemed to be very much on the negative/no effect side. I do think that good trials would be useful as it is so widely promoted, including by doctors who specialise in treating ME.
Perhaps the suggested 'no one size fits all' dosage problem could be largely overcome by having a baseline assessment for a month (e.g. including actimetry and daily pain scores), then a 3 month period when the dosage is adjusted to get to what the person thinks is the best dose in that time, and then three months at the 'individual best dose' with a treatment assessment in the last month. I'm not sure that the side effects are so predictable that they would break the blinding.
Do people who have tried LDN think that a trial like that would work? Obviously longer baseline, titration and stable dose treatment periods would be better if possible.
If given the choice of a trial of only LDN or Abilify, I think I'd vote for Abilify, because I think the potential side effects from Abilify are worse, and so it is more important to know if there is really an off-setting benefit.
At the risk of going off-topic, my PEM and that of many others here involves pain. I describe my pain when in PEM as like being crushed. It might be like the 'bone-crushing pain' that you get when you have the flu. ME/CFS is more than just fatigue. Mister Person, you might like to check out some of the discussion threads on PEM and other symptoms in
https://www.s4me.info/forums/symptoms-and-signs-discussions.197/ - you could ask your questions there.