The help center has links to discussion of the concept of good subjective questions vs bad subjective. In politics, it seems that many questions might well be subjective, but fit the parameters as good subjective, and specifically prompt answers long answers can be backed-up with facts.
The field of "political science" is intrinsically a social science, where subjective matters can be supported by empirical measurements without formal proof. But without proof they may not be answerable to the degree a hard science question can be.
So is putting a question on-hold because it seems opinion based reason to cut off answers to a well-scoped comparative question without permitting a several answers and/or refinement through discussion?
I guess another way of asking this might be can there be good opinion vs. bad opinion? I can easily see that such questions might be quickly protected to avoid trivial opinions, but answers containing how and why seem like they can add value in a community focused on social sciences.
Another relevant approach to this may be examining the politics.SE scorecard for graduation from Beta. I'd suggest we would move toward improving the number of answers per question by permitting more subjective questions. Take for example the clearly opinion-based question Why does partisanship trump concerns about hypocrisy with voters? There is no single answer. This question generated a substantial number of answers (currently 12), and the voting on answers pushed the democratically most correct to the top, without invalidating the less popular opinions.
The reader chooses how much 'noise' they want to consider by choosing when to stop reading from the sorted list.