Thank you Parvo.
Very Interesting. Examining this in detail makes you wonder how this team's earlier studies ever managed to get published.
Just one small point - and it is a small point in an otherwise excellent analysis. While the inclusion of psychological measures in the original study that provided the BMJ study cohort clearly highlights the motivations of the researchers and the underlying conceptual model, their inclusion and the analysis of these factors doesn't, in itself, necessarily strengthen the case for it not being an 'organic ME/CFS' cohort.
As you know a correlation, no matter how strong, does not necessarily mean causation, nor can one infer the 'direction' of the effect unless the study manipulates the IV. The researchers obviously hoped to demonstrate a correlation between psychological factors and the type and degree of symptoms with measures such a depression, locus of control, self efficacy etc being designated as the independent variables. However as the psychological and illness measures where taken concurrently, it would be equally legitimate to designate the measures of fatigue etc as the IV and the psychological measures as DV's. In fact it could be argued that there is greater face validity in positing that having a little understood, largely untreated and extremely disabling illness is extremely likely to have a negative impact on measures such as depression, feelings of control and self-efficacy in the expected direction.
In addition, as stated; they only found a weak correlation and had to conclude that much of the variability in the measures remained 'unexplained'.
Taken together, an equally strong case could be made that these people were in fact a 'well defined' cohort and that the study had failed because the factors driving the degree of illness were organic rather than psychological.
Not that I believe they were a well defined cohort, the other points made are more than suggestive that they were not. I just don't think this particular argument is the strongest.