He invents some concepts like "Lisp tradition" out of blue sky, based on a sample of three people, who he misrepresents and who aren't even serious Lisp programmers (except PG), and then fantasizes what he thinks the "traditions" (my ass, like old Lisp masters are convincing young adepts to not learn math or something) represent, so discussing this seriously is beside the point. Is the average C programmer better at mathematics? If anything, the books introducing Lisp programming go much more into modelling complicated domains, including mathematics, compare SICP or PAIP to K&R or your typical programming intro book nowadays.
Then, he seems to imply that focusing on recursion is somehow less mathematical. Recursion is a lot like mathematical induction, and inductive definitions are ubiquitous in mathematics. Also, a lot of mathematics is non-constructive and mathematicians aren't in general that interested in doing efficient computations, so it's more about knowing algorithms and numerical methods. So his big revelation reduces really to the fact that people doing numerical computations should learn numerical methods, or basically "know your domain", hardly a revelation. Or is it just a cry for programmers to learn more mathematics? But then the guys he criticizes have already done a much better job encouraging people to become interested in mathematics.