I didn't state that the document can't have audio or video or whatever. I am just sceptical when I see someone sending over the browser some wasm for singal processing or for rendering a video game. Don't get me wrong I think it's is useful but I am really worried that people are basically overcomplicate things with hacks on top of hacks instead of someone saying f#@k it we need a different approach.
My problem is that when I want to do some simple web app the industry standard is to have all these abstraction levels, this makes me think that maybe we are building on the wrong fundamentals. And people instead of refactoring the browser they make more abstractions/hacks on each other.