alt Hacker NewsThis is remarkably useful. Even ignoring the built-in commands (which are handy in their own right), I find the button's action being self-described in the html ("tell this element to do this") far more pleasant to read than the normal see button -> /document.getElementById("buttonID") -> scroll back up to the html to figure out what elements are referred to in the script.
To be fair, there's little stopping you from putting the button action in its onclick attribute.
I guess the stylistic choice of separating content, style, and interactivity eventually became a convention to keep JavaScript isolated from HTML, but nowadays with Tailwind and HTMX, it does seem like at least some developers want everything in HTML, for the strengths you mentioned.