Paul Krugman gave a presentation in Cambridge at a conference to celebrate the 75th anniversary of Keynes's General Theory of Employment, Interest and Money. He titled his presentation "Mr. Keynes and the Moderns" and in it he walks through the basic argument for what I will call "nuts and bolts Keynesianism" as opposed to either Chapter 12ers or the Samuelsonian synthesis. In the process, Krugman provides argument and evidence that challenges modern adherents of the "treasury view" and the "liquidationists". He argues that we are rehashing many of the debates once again that he thought were settled by Keynes back in the 1930s.
I like Krugman's article for a variety of reasons. For one, I am someone who believes the history of economic ideas is important for contemporary theory construction, so I have argued for what I term opportunistic readings of historical figures, see here and here. Krugman buys into that use of history of ideas when he states: "Keynes was a great man, but only a man, and our goal now is not to be faithful to his original intentions, but rather to enlist his help in dealing with the world as best as we can." This is Kenneth Boulding's idea of the "extended present." Second, I think he has presented the argument clearly and given specific empirical claims that must be addressed. Rather than wrestling with jello, Krugman provides a very clear argumentative target that is based on economics rather than political persuasion. It is Krugman as an economist not as a political pundit.
No doubt, there are "perspective" issues --- Krugman's dichotomization of the Chapter 12ers and the Keynes as equilibrium theorist and the focus on the critique of Say's Law, does reveal an inability to deal with Say's Law as separate from Say's identity and as a process theory as opposed to an equilibrium condition. The classics on this would be W. H. Hutt's A Rehabilitation of Say's Law and the classic article on this would be Axel Leijonhufvud's "Say's Principle: What It Means and What It Doesn't Mean", reprinted in Information and Coordination. Bottom line: Say didn't confuse an accounting identity with causal relationships, and neither did Hutt or Leijonhufvud. Say's critique of Malthus is as relevant today as it was then and simply amounts to a reaffirming that there are economic forces at work in an economy. A huge mistake in perspective occurs when thinkers (and policy makers following them) assert either that we must deny these forces operational power (cut the story short so to speak as happens when agent rationality or relative price adjustments are denied) or assume they are represented by instantaneous adjustments so at any point in time everything is fully worked out.
But for sake of argument, let's put the differences in perspective aside. Instead, focus on the causal relationships that Krugman identifies and the evidence he employs to illustrate his position.