I agree with you that the close reason is not well chosen. It seems highly likely to me that the first close voter selected "primarily opinion based" because it was the prefab close reason that came closest to the actual reason the voter had for wanting to close the question. If the voter had taken the time to write a custom close reason, it probably would have said something along the lines of "I am voting to close this question as off topic because it is a request for writing advice."
The bigger issue that your Meta post raises is whether using the closest multiple-choice close reason when none really fits is a sound practice. Obviously the benefit to the close voter is that it helps move an objectionable question nearer to being closed with a minimum of effort on the close voter’s part. The harm is primarily to the person who posted the question, but also to EL&U's goal of generating better questions. After all, the theory behind identifying a reason for closure (as I understand it) is to educate the person who posted the faulty question—and others—as to what is wrong with it. If that weren’t the point, Stack Exchange might as well permit unexplained, drive-by close voting, just as it does unexplained, drive-by downvoting.
If someone posts an answer that is objectionable because it is green, and the multiple-choice close reasons available are red, yellow, blue, orange, and purple, a close voter might argue that selecting “I am voting to close this question because it is blue” or “I am voting to close this question because it is yellow” is the most reasonable choice—but that rationale ignores the option of writing a customized answer that identifies what is actually wrong with the question. And of course, if you don't accurately explain a question’s failings, you don’t enlighten the person who posted the question about the question’s real defects.
This leads to a second problem with choosing an inaccurate prefab close reason: stringing the questioner along. I see this happen most often with the “show research” close reason. Someone asks a writing advice question or a homework question or too-localized-to-be-of-interest-to-anyone-else question, and close voters respond by voting to close on grounds that the questioner didn’t show any results of personal research. Every now and then, in response to such a closure, a questioner will go back and say “I looked in reference books and couldn’t find an answer to my question” and ask that the question be reopened—and then one of two things happens: reviewers in the Reopen Votes queue decline to reopen the question because they implicitly recognize that the actual defectiveness of the original question was not cured by showing research, or they reopen the question and then they or other reviewers vote to reclose the question on other grounds.
Far too often, we review-queue voters are sloppy, inaccurate, and lazy in our application of close reasons. We would show more respect for everyone concerned if we held ourselves to the standard of identifying what is actually wrong with a faulty question, rather than taking the easy route of selecting the most nearly applicable multiple-choice option, even when it isn’t particularly near. At the very least, close voters who don’t want to bother with providing an on-point justification for their close vote might consider selecting ‘off-topic because...’ > ‘Other (add a comment explaining what is wrong)’ > ‘I'm voting to close this question as off-topic because it’s bad.’ They’ll spend only slightly more time voting to close in this way than they do now, and they won’t mislead OPs with spurious nearest-prefab-choice close reasons.