"you can easily fix most Bootstrap bugs by overriding it with some custom CSS" vs. "you can easily fix most [imperative/OO/functional programming language's library] bugs by overriding it with some custom [code]"
and
"just not use that CSS class" vs. "just not use that [method/function/class]"
An contrived example: let's say an unfixed Bootstrap 3 bug is that there is a formatting error when rendering "div.jumbotron > h1 > span.label > small". You could replace the last small with you own css class and be done.
Contrast that to a bug in Angular, for example, that enabled a XSS bug. You'd want to upgrade Angular, instead of manually patching or adding a workaround to all your forms.
That's what I was thinking, at least.
Then why say that you're ending the support for Bootstrap 3? Clearly something IS not going to be done from this point onward, and it's that thing which is worrying people.
What they are not "ending" at all is "We still recommend v3 for production and believe it to be stable".
The complication here is the two meanings of "support" of "recommendation to use for your production efforts" versus "we are actively helping a tiny fraction of users with long-tail bug reports".
It's a pile of CSS and a few scripts to show and hide things. And it has had three years of effort put in to make sure everything works mostly as it should. That's not going to change.
We're just not going to get anything new. It was fine for production use yesterday. What has changed that makes it less fine today?