I think you missed the point of Jonathan's article as well. It wasn't even about COBOL. The point is that innovation needs to happen all the way around, not just in languages. The industries that still make use of COBOL are stuck in the 50's, and the language is just a small part of that issue.