> Second-guessing their doctors and refusing to vaccinate their kids after reading stuff written by people who know exactly zip about medicine.
This claim was originally by an expert doctor.
> Which is why articles and books are subjected to “peer review” and not to “everyone review,”
Peer review struggles to correct inaccuracies and by failing to share the underlying data behind the assertions made, and especially in Tom Nichols' area of expertise (public policy) the data is the only way of invalidating claims (as opposed to abstract mathematical concepts, like machine learning). Further more, most peer review journals are behind paywalls of some kind.
Contrast this to the accuracy of Wikipedia (where, yes peers review the knowledge, but so can the general public) or the increasing support behind open peer review (http://en.wikipedia.org/wiki/Open_peer_review) and the open research movement in general (http://en.wikipedia.org/wiki/Open_research) it is at least questionable to argue that closed peer review implies a standard of truth.
> but don’t tell that to someone hectoring you about the how things really work in Moscow or Beijing or Washington.
It is funny he brings this up because Beijing is rife with academic dishonesty to the point where if the exact steps / data to reproduce aren't explicitly mentioned in the paper, I don't even read it.
> [A large, large amount of opinionated elitism ending with a call for the return of the days of gatekeepers]
Gatekeepers impede scientific inquiry. It wasn't until 1957 that we seriously started examining human sexual response. This is 100% due to gatekeepers. By advancing the case for open data, open methods, and open journals we may suffer more fools' opinion's on global warming, but it is far more worth it.
The reason I suspect that this author is especially angry is that public policy is politics in declarative form. "We will use one school of 500 kids vs five schools of 100 kids" has far reaching consequences that spill into the political sphere. The reason public policy is hard to take a seriously as a scientific endeavour is that it isn't often provable, although with open data would certainly improve it.