I love this article! Surprised that there's not much discussion ensuing from it. I think the quest for more performant tooling is a worthy thing in itself, as the Python community has had great tooling, but not so much performant tooling, and things like this have spillover effects on other projects.