I asked a question earlier today that I was half-expecting to be closed, largely because I wanted to test out whether it would be. That is was closed within only two hours indicates something highly problematic about our site: by the rules we're currently following, there's no good way to ask certain types of questions.
So let's suppose that I'm a new UX designer looking to buy some new wireframing software. A practical, concrete, real-world need, that most UX designers face at some point or another.
Obviously, I can't just ask "what's a good wireframing application?" because opinions would be too subjective, because people's attachment to their own products would likely spark extended debate, and because the most popular answers would grow out-dated over time. This is too narrow a way of phrasing the question.
However, if I instead broaden the question and ask "how should I evaluate wireframing software?" or the slightly more concrete "what features should I value most when choosing wireframing software?" the question is still closed. Instead of being too narrow, it becomes a request for subjective responses.
Between the two extrema, I could ask other questions, such as "what are the best practices for converting wireframes to prototypes?" or "should the styles of a wireframe look intentionally unpolished?" but such questions would fail to address my underlying information need because there might be important features of wireframing software that I would completely overlook (and thus not know to ask about). Such questions could be useful follow-ups to my first question, but don't stand as well on their own.
Questions asking how to evaluate UX software are a good fit for this site and address a common, answerable information need. Considering the good subjective, bad subjective test, such questions fall under the category of good subjective. Yes, let's remove bad (overly subjective) answers, but let's not close the questions just because it's possible to give a bad answer.
Good answers to such questions will come from "experts," which is to say people who have already used wireframing software and know what makes it good or bad. These answers can be supported by concrete evidence, stating why particular features are essential to have in whatever product you use. Such answers will be highly useful to anyone who comes to the site wanting to know how to pick wireframing software: a practical, concrete, real-world answer for their practical concrete real-world need.
There may be a few different viewpoints, but that's the beauty of this site's structure, isn't it? We vote the best answer to the top. As long as it doesn't turn into a flamewar, having two or three different good answers harms no one. Even for the most objective of questions, there can be multiple, subjective answers, each providing value to the asker.
We have plenty of expert users on this site. Why not use their expertise to do something more than just look at trivial details of interface design ("should I lay out the form Way A or Way B?") and ask some bigger questions about the issues that UX designers face?