Casma Shalizi has written a delightful article, except maybe for the last point:
Points 1, 2, 3, 4 and 5 simply state the obvious, but it gets real interesting when in point 5 Casma can't see why errors in macro-level theory "should always be the preferred option in economics". Very good! Maybe there are micro-level errors after all! Accordingly, "maybe the simplification of always solving for the equilibrium is wrong". Perfect! Mainstream economics long-term equilibrium equations assumption of homo economicus is a perfect example of a micro-level error, which "accumulates when one goes to the level of whole economies". Wonderful! Finally, Casma finds an analogy that "classical physics predicts that matter should be unstable". Another excellent example of greedy reductionism based on primitive (classical) micro-level models!
Then in point 6 Casma continues to state the obvious, Austrians do usually develop their theory post-hoc. And that is correct, after all, also virtually all mathematical proofs have been developed post-hoc. But you better have correct micro-level theory first!
Finally, in point 7, Casma believes that macro-level causal model is easier to develop than micro-level one. Certainly! Greedy reductionism example! But why should something be "more suitable for policy-making" just because it is easier? Beats me...