Related to the last post, I wanted to bring your attention to LK's first post on Lee's Post Keynesian Price Theory, which draws a lot on Means. Means was not a Post Keynesian, he was an institutionalist, but his administrative pricing work obviously holds interest for Post Keynesians that drop optimization, use mark-up pricing, and are seeking out justifications for the trade.
As you might guess, I'd love to see a bit of both. Optimization is a powerful tool that explains human behavior quite well. It is flexible and requires relatively weak assumptions. It is, however, a little one dimensional. So I do think we need a lot more institutionalism in economics than we currently have. We need more field work and more economists that write like sociologists. But we don't need that because there's something fallacious about optimization. We need it because although optimization really gets at some essential issues thing are more complex than that.