I wonder if EMA (Elevator-Monopoly-Adele therapy), or something like it, would be a useful control in some of these studies, particularly in children. The idea being to compare the response to a nonsensical intervention to the proposed intervention. Quite possibly, both would give equivalent results. Then you'd have to choose between two alternatives: that EMA therapy actually works, or that patients tend to endorse any intervention they participate in, so long as it seems minimally plausible.
In 1987, Dr. Steven Straus of the NIH said that, "[CFS] patients will commonly feel better - no matter what you give them," implying that the illness is imagined. But that statement can also apply to many people with physical illnesses. It's called "faith healing" and patients with all sorts of indisputable illnesses may "commonly feel better," after receiving it. However, that doesn't mean that there's been any improvement in their underlying disease.
I guess I'm just suggesting that maybe there should be a "faith healing" control on some of these studies.