Let me start by trying to boil down the issue with (primarily) opinion based questions: The answers to such questions can't be properly evaluated on their quality. That means (opposing) answers will be (up)voted for their popularity instead.
We have some means of identifying quality, most relevant here is being supported by experience based expertise (aka Good Subjective). Then, when evaluating a given question, there are two questions to ask:
Can there be a well founded and identifiable as good answer to this?
Is it inviting answers which aren't that?
Answering that will come down to the curator's (ie. close votes) judgement. There are simply too many things that can be one way or the other to quantify it a lot better, at least as a way to cover for all questions.†
Those two tester questions can be true and false independently, but should generally be edited so that 1. is obviously clear and 2. is false. If the needed edit is obvious, such as removing a "what are your ideas about this?" or some-such to remove 2., please just do that. Otherwise it'll need to be closed until those are clear.
And as a final note, sometimes we can't really tell what answers a question will get until they're in. Sometimes we might have concerns and give the question the benefit of the doubt and a trial by fire (maybe with a stern reminder about citing experience), or a question which looked fine will have answers which aren't. Either way, questions are sometimes closed based on their answers demonstrating them opinion based. Obviously that's a lot messier, and so there is a strong preference towards getting them fixed first.
†: You can still have question types where it becomes easier to identify good marks and issues, though adding specific rules is generally not gonna save a question type. See game rec.
