I don't. Website updates are mostly changes to the content once it's gone live, and that shouldn't require a developer. Having a CMS in the backend and a process to rebuild a static website from the content when it's changed lowers the cost of keeping things up to date, and improves the web by making websites stay relevant.
What the code looks like when a dev is working on it, or when it's sitting in a repo, it's mostly irrelevant. The important things are that the site owner can keep the content accurate and up to date, and the HTML and CSS that's delivered to the users is high quality and efficient.
Unless every page is entirely unique, a web site consisting of raw HTML pages has a lot of repetition / redundancy, and coordinated updates become repetitive, too. Authoring, too.
The upside, of course, is that there's literally nothing to go wrong in the pipeline, because there is none beside the upload operation.
I can imagine that sites like https://ciechanow.ski where every page appears custom-made could be managed like that.
Again, I don't like taking sides. Just searching for absolute truth. We will never get there, but we can try.
ref to your comment here: https://news.ycombinator.com/item?id=34019912