The blog post clarified it's about maintainers of critical packages, not all contributors. This could be limited to packages with just one or two maintainers, especially newer ones. And they could remain somewhat anonymous, providing their information to trusted third parties only. If some maintainers don’t accept even this, their commits could be put into some special queue that requires additional people to sign off on them before they get accepted downstream. It's not a complete fix, but it should help.