We can draw the line at a different degree of opinion-ness, but we should still have this close reason. Consider some possible questions from someone thinking to buy an X and considering Y feature in it:
These three really are not good questions. They don't help other people later, and they don't much help the asker now unless the answerer happens to provide a ton of detail that wasn't explicitly asked for.
- what X should I buy?
- are X's with Y worth the extra money?
- should my X have a Y?
These three are better. They may still involve opinion (how important is a window in a tent, how heavy is too heavy for a backpacker, how cold or warm is too cold or warm) but they have some objective component and they aren't asking us to make a decision for the asker.
- what difference does it make if my X has a Y?
- what are the advantages of a Y for my X?
- How do I decide about a Y for my X?
We can close the first three as primarily opinion based and encourage them to be edited into something more like the last three.