I personally find that people who admit their mistakes more trustworthy because it shows they can at least learn from them. In this case, the jury is very much still out on whether BTC will succeed as a currency, and not as a protocol that's useful for some other thing (or many other things), which seems to be the crux of Krugman's argument.