FWIW, I don't think it really does to treat his opinions on typing as a blanket, universal, prescriptive statement that one can or should agree or disagree with according to one's own blanket, universal, prescriptive position. He generally doesn't come out and say this explicitly, but, in his talks, he's always speaking to a certain audience, about a certain class of problems.
You couldn't come up with a better pair of languages than Clojure and Rust for illustrating this sort of thing. I like them both (as far as my extremely limited experience in each will allow, anyway). But, if a language like Clojure is a candidate for the thing I'm working on, then I would never have dreamed of picking a language like Rust. And vice versa.
To add to that and concur that Rich generally doesn't have such un-nuanced strong opinions without context, see his talk Effective Programs. He explains what Clojure is good for and isn't good for, and says more clearly it so depends on what you're doing.
He doesn't organize the Clojure conferences, but I'd imagine that if he didn't see any value in any context for static typing, there wouldn't have been a talk about Typed Racket and static typing researchers as keynote speakers, etc.
And FWIW, spec is a library trying to come as close as possible to the imaginary midpoint of static and dynamic typing, but in a Clojure-y way.
You couldn't come up with a better pair of languages than Clojure and Rust for illustrating this sort of thing. I like them both (as far as my extremely limited experience in each will allow, anyway). But, if a language like Clojure is a candidate for the thing I'm working on, then I would never have dreamed of picking a language like Rust. And vice versa.