> Their inability to come up with accurate priors is completely lost on many of the folks who follow this kind of thinking.

Mis-specified/faulty priors are absolutely a possible pitfall of Bayes, but Jeffreys Priors are a means to circumvent this: https://en.wikipedia.org/wiki/Jeffreys_prior

Another thing not mentioned in the article is that Bayes DOES converge to the frequentist result in the limit of big data.