I would like to summarize some of rlc's excellent points, mostly for myself, but also for anyone else who, like me, often struggle to get through multiple large posts at once and are tempted to gloss over them instead.
[subject to editorialization and further modification]
The recent study on XMRV/pMLV, irrespective of the quality of the virology, did not do enough to exclude known medical conditions from the ME/CFS cohort (the list of tests performed can be found in Table 1 of the paper:
http://mbio.asm.org/content/3/5/e00266-12/T1.expansion.html). Unfortunately, this is a very common problem in the research literature on ME/CFS, which is primarily comprised of small studies without adequate confirmation or follow-up attempts at replication.
There are many different and conflicting criteria for ME and/or CFS, which makes it difficult to define it for research and has lead to mixed cohorts. A start would be thoroughly excluding all patients with other diseases by extensively testing the study candidates (although it is also important not to accidentally exclude those with clinical characteristics previously associated with ME/CFS eg OI). Without taking all these cautions, efforts into projects such as the CFI will be wasted. We also cannot assume that when a research group uses a stated criteria, that all the tests suggested by those criteria were actually done.
It is difficult to judge a "top ME/CFS expert", because there is no reliable standards to compare with, and little established/replicated science on the issues of testing and treatment. Anyone can present themselves as an expert or be accepted as such, by being familiar with much of the (unreplicated) research and demonstrating great sympathy towards patients, but are often very expensive and not necessarily effective doctors.
