Wow, this is totally counter to my experience. On my team we explicitly frame PR review as an opportunity to pick up slack of other team members when necessary.

Wow, this is totally counter to my experience. On my team we explicitly frame PR review as an opportunity to pick up slack of other team members when necessary. We trust each other but, put simply, "shit happens" and people have bad days, miss things, aren't as familiar with parts of the code base as others, etc. PR review, even for experienced and senior engineers, is us admitting that humans are flawed and we should account for that in shipping software. It has also been invaluable for onboarding junior engineers and giving them an avenue to contribute in a controlled manner that also allows us to each them in an asynchronously. They can review the notes we leave on their PRs and learn at their own pace. I guess the alternative is to do pair programming 100% of the time, ditch PR review and commit directly to master? It's certainly an interesting proposal.