Yeah that is interesting - especially the stuff about central banks as monasteries - I think that is absolutely true. The history of macroeconomics over the last few decades would look very different if based on research in central banks and treasuries, etc, rather than the academic journals. I can't understand how smart people take something like real business cycle theory seriously.
He exaggerates how hard things have been for the 'saltwater economists' though... Krugman, DeLong et al were hardly crying in the wilderness in the 1990s and 2000s - they were absolutely Establishment. And it wasn't only Clintonites that were Keynesians - Mankiw is a New Keynesian too. I think the major ideological enemy of the technocrats is not so much the freshwater economists as plain old know-nothing conservatism - the 'commonsense' view of government debt and so on.
As for the cognitive dissonance between macro and micro - Krugman advocates learning to live with it. I would say rather that macroeconomists should have the courage of their convictions. The problem is not that macroeconomics lacks microfoundations, but that microeconomics lacks macrofoundations... and a serious treatment of time and uncertainty.
Kevin D. Hoover has written some great stuff on the history of macroeonomics, and on methodology, from a still fairly mainstream perspective: http://econ.duke.edu/~kdh9/
Mike Beggs