More importantly, Krugman actually admits when he is wrong based on new data, as he has done on several occasions in his blog and the NYT. To me, that is the mark of a scientist, one who is able to be swayed by data.
Very few of the Austrians and Chicago types seem to admit to being wrong, and often, contradictory data is simply waved away as conspiratorial (inflation, econometrics). The whole notion of praxeology turns their conclusions into untestable pseudo-science. It takes on the air of a religion rather than geeky engineers willing to hold their ideas to fire of real world tests.
There are two issues at play here: the deep desire of technolibertarians for "free money" and decentralization, and moral arguments surrounding the function of the state and application of force, and the real world where their predictions often fail. It's possible to love Bitcoin, but also admit Krugman is right about it.
Bitcoin is a magnificent creation of cryptography. In fact, the entire field of modern cryptography, from one way functions, zero knowledge proofs, oblivious transfer, to high level schemes involving money, voting, homomorphic encryption, et al, is staggering beautiful. When I first read of Broadcast Encryption, despite its use in DRM, my mind was blown. I love all this stuff, and have a deep affinity for subversive systems.
But Bitcoin isn't the anarchocapitalist utopia enabler that converts the world into Galt's Gulch, because the deep beliefs about how a totally free economy would work in practice don't have to be any more true than the beliefs about how anarcho-syndicalist communities would work when scaled up. The real world is complex, and a mixed set of strategies is more adaptive and stable than a set of purist ones.