Yeah, the "random guesswork" part feels spot on. I had a paper flagged once and it was a total mess. I just ran it through a few different checkers and they all gave me wildly different results. Ended up having to just trust my own work and the feedback from my professor. Those tools seem to create more stress than they solve.