I have a few thoughts -
1. First, I'm not sure I even entirely agree with the premise of the question. Sure there are lots of Keynesian strawmen out there, but that's the case for a lot of economic theories. RBC and EMH folks have no sympathy at all for the Keynesians right now when it comes to strawmen version of their ideas. Are Keynesians especially embattled in this respect? I'm not so sure.
2. Ideology. Unfortunately it's always a part of the discussion in economics. I don't think it has to be or that it should be, but it plays a role. There are a lot of people that address strawman Keynesianism because they simply don't care about good economic analysis.
3. The Neoclassical Synthesis. Here I have to agree with the Post Keynesians in that this downplayed a major facet of the General Theory - namely the important distinction between the interest rate and the marginal efficiency of capital. A lot of the neoclassical synthesis was quite productive. I'm not universally a critic. But it turned Keynesianism into something else. Keynesianism is a family of theories, which I think makes some people think it's incoherent.
4. Poor teaching. Unfortunately, it's very easy to get "Y=C+I+G so if you boost G you boost Y" in a lower level macro class. Since lots of people don't take more than these lower level courses, they find themselves in the blogosphere or a policy debate with that image of Keynesianism in their mind.
5. New Keynesianism. Again, like the neoclassical synthesis I think the New Keynesians bring a lot of fantastic stuff to the table - but, they ironically revive Pigovianism and frictional stuff and get that tied back in with Keynesianism. Keynes noted frictions and price stickiness. There's nothing non-Keynesian about it. But the dominance of the New Keynesians and the oversimplification of New Keynesianism as price stickiness obscures what made Keynes unique.
6. Keynes's writing style. When you write about pyramids and dirges for the dead and burying bank notes reasonable people that appreciate a good writer see it for what it is - some very clever illustrations of fundamental principles. People not interested in really engaging Keynes completely miss his points. One of my all time favorite examples of this is a piece that Paul Canotr wrote for mises.org on the pyramid passage. It was one of the first Mises posts I ever read, actually. The man is stunningly clueless about what Keynes was saying. It's especially a shame to see people make this mistake, because Keynes really has a way with the English language - he's one of those writers that is always a pleasure to read, regardless of what he's talking about.
7. Accounting identities. People don't understand them. This is something where sophomoric critics can be the worst. People who know just enough economics to know some of the basic relations and be interested in talking about it invoke some of the most banal, erroneous claims as if they were deep insights. "You can't spend more than you take in therefore the government can't keep running deficits forever", "if you consume more you have to invest less". Etc., etc.
8. The presumption of ideological orthogonality. The assumption that "since I believe X, and you and I disagree, you must believe not-X".
I think Keynes was fairly clear about what he thought, and when he wasn't sure what he thought he went to great lengths to say he wasn't sure and that we'd have to figure it out through experience. He didn't say all the things that Keynesians say today, but that's OK I think. With the exception of where Hicks and Hansen took the determination of the interest rate, nothing modern Keynesians say today strike me as being dramatically opposed to what Keynes could have thought. If he had the luxury of living until today, I imagine he'd take up a lot of modern Keynesianism. So I think it's silly to contrast "Keynesian economics" and "the economics of Keynes" too strenuously as a debating point (although it's certainly interesting intellectual history.