Source for anecdotal data = hearsay from a stockblogger with no background in science.
Research data in question is less important than hearsay about the forementioned research data, (i.e. opinions from those with conflicts of interest.)
Interpretations of poorly generated data about the research is less relevant than the opinion of the stockblogger, (i.e. only believe the data interpretations when positive.)
Maybe you should try to explain that post again. Please be more clear because this is a little confusing.