Definitely not exclusive, though I believe they are related enough to not tank one for the other. If a review is 2 seconds and done, then the circle for figuring out the mvp slop app is pretty slim.
You can imagine with the AI models we have now, many of the stuff reviewers do could be automated to a large extent. They can also determine through binary diffing that functionality changes may be limited etc. Many things can be done to eliminate majority of the queue for example.
Some stuff the reviewers do could easily be automated. I’m off the mindset everything the reviewers do shouldn’t be automated however, as llms do hallucinate and are still susceptible to ghost prompts.
At the infinite app review scale (some apps malicious, illegal, rip-offs), extra time is inevitable.