It's true that there are not enough "no"s. And at the same time there are also too many "no"s. E.g., many PRs are not a PR because one wants that change in the

It's true that there are not enough "no"s. And at the same time there are also too many "no"s. E.g., many PRs are not a PR because one wants that change in the code, but because one has a problem that needs solving and I thought about helping the developers instead of letting them do all the work. If you just close it without recognizing it in any way then the problem stays still unfixed, and motivate people to do things you won't like, like patching your project just in their company's environment and thereby creating a hidden fork. In the end I think the question is not about "yes" or "no" but about what's the actual problem and not making the easiest choice but trying to make the best.