The wrong (not cynical enough!) criticism of Bayesianism

In her important book, Science. Policy, and the Value-free Ideal, Heather Douglas (a former most-underrated) has an important footnote:

...the Bayesian framework, has yet to be shown useful in real world contexts where both likelihoods and priors are disputed. At best, it serves as general guidance for how one's views should shift in the face of evidenec. When both likelihoods and priors are disputed, abundant evidence may still never produce a convergence of probability. (p. 183 n 15)

Let's leave aside (for the sake of argument) recent advances within Bayesianism that try to deal with model choice under uncertainty. What this criticism fails to note is that what Douglas views as a problem is, in fact, the main de facto reason for the widespread use of the Bayesian framework in policy circumstances: it is precisely its (very useful!) flexible ability (one can play around with the priors) that permits one to justify scientifically and "technically" almost any (bureaucratically/politically wished for) course of action. The dirty secret about Bayesianism is, thus, that its very flexibility makes it the perfect policy making and grant-writing  tool!

Update: For more technical criticisms of Bayesianism, see John Norton's very useful papers.