Specifically, Williamson is right about the state of macroeconomics and Krugman is right about IS-LM.
And Williamson is just being cantankerous and missing the point when he claims that IS-LM can't explain things it wasn't designed to explain. The point is that on one of the crucial questions that we needed to get right, IS-LM does a great job describing our predicament - a lot better than what some people said who long ago abandoned IS-LM as some kind of joke or dead end. Of course the conversation doesn't end there, even in evaluating the questions IS-LM does speak to. And of course there are a lot of important questions it doesn't speak to! The point is this: if economic policymakers knew no macroeconomics but IS-LM and they were fully convinced of the value of IS-LM we would be in a better place than we are today. That's all.
Now - when I say that Williamson is right about the state of macroeconomics, I should clarify that that means I think Krugman is wrong. He's way too down on mainstream macro... reminds me of a few other people I know.