Well, popular (or increasingly mainstream) results get more grants, peer acceptance and media coverage, so there's always that.
Up to the eighties those same journals would have no qualms "proving" that homosexuality is harmful, and in fact there were tons of studies to that effect.
(It works the same, but with an inverse sign, for those scientists going for the conservative audience and its, increasingly fewer, and more marginalised, grants).
So it could be more that the tide has turned, and a push to get a cute result to publish and get some publicity and good will, than an actual causation.
Better assign the same, high, probability in all fields where there are humans involved.
I assign a similarly low probability to the idea that the refuting author's self interest or fallibility caused him to publish an opinion that is at odds with the current mission of the organization that published it as I do to the probability that the original author is at odds with the purposes of the institution that published his research.
Any statistical study can be criticized as 'only being likely, not being certain.' Briggs does not provide contrary epidemiological evidence. His critique is based on the possibility of alternative explanation inherent in all statistical studies. At best there's a failure to understand the epistemological character of empirical knowledge.