How to "fix the internet"? Stop advocating for JS-heavy/required sites (including SPAs) whose use requires browsers that depend on a massively complex specification that can only be implemented fully by a huge organisation, and advocate the subset of HTML that would be usable from nearly any browser --- even one that any reasonably skilled developer could write him/herself --- and the highly accessible and user-friendly "internet as a hyperlinked web of documents" which it was envisioned to be.
To try to "fix the internet" in a commercial way just feels a little awkward.
"Stop advocating for JS-heavy/required sites (including SPAs) whose use requires browsers that depend on a massively complex specification that can only be implemented fully by a huge organisation, and advocate the subset of HTML that would be usable from nearly any browser --- even one that any reasonably skilled developer could write him/herself" ...... Yes, I agree. (There is occasionally a use for such things, but it should be avoided as much as possible.)
Furthermore, web browsers need to be fixed too, to give the user enough ropes to hang themself and also a few more just in case, and to block commands coming from the server.
Furthermore, web browser should be deprecated for many things. There is such thing as Telnet, SSH, NNTP, IRC, etc, that can be in use, that should be encouraged to be in use.
Furthermore, you could host stuff on your own computer if necessary.
I also like the idea (mentioned in another comment) to call much of the mess as "internet pollution". Yes, these ads, scripts, etc, wastes bandwidth, so it is "internet pollution".
Nothing about that is inherently bad. It leads to bad behaviour, but in and of itself it's not bad. Also, saying that only a big corporation can provide a browser is both the nature of things and a lie. There are hugely complex programs built by "the community" (VLC to give an example).
I want a better internet but I don't want to go back to the past.
There are hugely complex programs built by "the community" (VLC to give an example).
The complexity of a browser and the complexity of a media player are extremely different. To make a huge generalisation, the latter is far more maths-heavy and less interactive. I have played around with writing decoders for a few codecs, and to a first approximation, they are very similar, so it's not hard to see how something like VLC was created. Media formats are also relatively stable, it's not like a new one comes out every few months.
On the other hand, the modern browser has many different components that a very wide breadth of knowledge is necessary, and on top of that, companies like Google are paying a great many people to do what appears to simply be churning the "standards" and continuously increasing their complexity, making it harder for others to even keep up. They invented the absurd oxymoron "living standard". Their common euphemism for that is "pushing the web forward".
The great diversity of media players vs. the effective monopoly of browser(s) already shows this huge difference.
To try to "fix the internet" in a commercial way just feels a little awkward.