The thread Examples of poor review audits contains other such examples. Yes, the review audits are picked automatically, based on the data like score, post history (close/delete events) and maybe the number of views, who knows. The algorithm is not disclosed.
Naturally, decisions to close or not to close are not a function of question score. One can sincerely believe that a popular question should have been closed, and get blamed for not paying attention. This is neither more nor less than a little annoyance and loss of time.
For Close review specifically, I tend to open most questions in a new tab and hit Skip in review; then vote on the question directly. (This process can be streamlined by bookmarklets). E.g., "known bad" audit questions are already deleted, so I don't have to waste time picking a close reason for them. This also allows me to review more than $20$ questions from Close queue per day. (Cue meta outrage about exploiting a loophole.)