Earlier quoted context omitted.
No, because once you download the js you don't redownload it on every page request. When you want to update contents you get the new data to be displayed. This is how any other gui application does server communication. The traditional page based responses re-send all the html content over again. If the client can cache it then it doesn't have to be, but caching something like a git frontend does not work with page b…
When most of the web looks like this: https://news.ycombinator.com/item?id=17655316 I'm happy about Drew's decision to not include JS.
This is 2018. You don't have a 56K modem, I don't have one. Those that do will benefit from smaller data requests after the initial download. Holding back the real issues the web can solve over an arbitrary problem like that is damaging to what can be achieved.
The web is a software platform, and unless you want to equally complain about applications in C that go over 1MB, you're making mountains out of mole hills.