Although your response is a bit clever, I think you capture the actual dynamic, which the paper (https://www.nature.com/articles/s41467-023-38626-y) tries to obscure by avoiding the comparisons (time taken for each correctly answered question vs incorrect) needed to make it obvious.
On a multiple choice test, does it really matter if they filter to only correct answers? Guessing will still result in fast answers, and be correct 25% of the time.
