First, it is trivially true that degrees in a subject are not a good indicator of having really learned it. As Biker pointed out, that is true in both CS and literary criticism.

Secondly I think it is true that mastery of the current research in either, particularly of the current fads, is both not necessary to producing the product (be it literature or programs), and I think that a survey of noted professionals of either topic would back me up.

However, and this is a big however, I suspect that it is hard to impossible to find a good author who has not thought long and hard about the writing process, and I have never to the best of my knowledge met or heard of a good programmer who has not thought long and hard about the programming process. And that thinking inevitably includes exposure to and thinking about the academic study of the topic in question (programming or literature).

So I agree with the thesis in several ways. I don't think that degrees indicate knowledge, I think that academic fashion can and does depart from what is useful for the practice, but practitioners are well advised to understand the fundamentals of the academic theory.