It is about a certain class of interactive problems where you know that you can solve it in $$$n$$$ queries but instead the author wants you to solve it in $$$n - 1$$$ queries, I feel those kind of interactive problems are straight up bullshit.

From what I've observed, these problems follow a similar pattern: the natural solution is pretty straightforward, but the constraints doesn't benefit the solution and rather add unnecessary complexity.

This feels like over-engineering to me instead of encouraging algorithmic thinking. In the end, these problems waste everyone's time with negligible value addition.