1. Some of us just have an interest in history of thought and in Keynes himself, and that's OK. I'm one of those people.
2. Many people conflate Keynes and Keynesians, and so Keynesians are smeared by criticizing Keynes or vice versa - so setting people straight often requires going back to Keynes himself. If that's going to be what they do, that's how we'll have to respond.
3. A lot of the distinction between "Keynes" and "Keynesians" is very superficial or technical anyway. Economics has progressed a lot since Keynes - it's gotten a lot more technical. But Hicks and Samuelson were basically hitting on the same points Keynes did and despite Krugman's frustration with DSGE models and New Keynesianism, this work incorporates processes that function like the old Keynesian IS and LM curves that he likes. Krugman should be more like Mark Thoma and George Evans in this regard and realize that while it's true New Keynesians can slip into talking like pre-Keynesian Pigovians, there's a ton of common ground. If Keynes is still providing the fundamental foundation for modern Keynesians (I think he is) in the sense that he is providing the big picture and the motivating ideas, then it's more reasonable to stand up for Keynes against nonsense criticism.
You know what I think's going on? I think Krugman knows he's usually the one tagged as being "shrill" and he's taking this opportunity to be the guy that's playing it cool :)
Krugman also writes this: "modern Keynesianism is to be understood through the views of modern Keynesians, not by hunting through the original works for hidden meanings", and I think this is very true. I talk a lot about this sort of stuff on the blog because it's easier to talk about than other things, and it's a personal interest of mine. But you don't do economic science by browsing through dusty books. I'm figuring out a rough idea of what I'm going to write my dissertation on (a little ahead of the game, I know) - and I'm mostly thinking through work that Robert Shimer published just last year, and some stuff by Pissarides and Mortensen that they just got the Nobel Prize for last year (it's not brand new material, but the point is it is recognized as a very modern paradigm in the science, relative to the history of the discipline). The farthest back I'm even thinking about for my dissertation is some labor contract literature from the 1980s.