In an ideal world, yes, we would patiently sit down with each poster, discuss what in particular is unclear about their question and guide them to a clear question, which we would then proceed to answer.
Unfortunately, it appears to me like the supply of unclarity is unlimited.
To be honest, we get questions that are so confused that I wouldn't even know where to start asking for clarifications. And my bitter experience is that sometimes when you do ask, the "clarifications" confuse me even more. In addition, sometimes the OP seems to have way more time to write up obfuscating "clarifications" than I have to deal with them.
And if I wrote the original request for clarification, I feel like I should deal with the follow-up discussion, which can take way more time than I have. Rarely do I request clarification, the OP provides it, and a different user answers the now-clarified question. (I may need to work on this notion of mine that commenting on a question confers some kind of "ownership" of it on me.)
So I personally will continue working CV as I have been doing: if the original question gives me hope that it can be clarified reasonably quickly (as in: the OP obviously invested some effort to create a decent question, as, e.g., here), then yes, I will comment and ask for clarification. Questions where I fear a fruitless time sink I will continue to close-vote without explanations (no examples here, but yes, I could provide them). For time is limited.
Based on a completely non-scientific survey of my close votes queue, it seems like others on this site proceed in the same way.
Yes, this is a classification problem, and Peter notes that some questions
are completely unclear and it's fairly obvious that they need to be closed.
However, there is a continuum here. A question may cross the subjective threshold to complete unclarity for one user, but not for another. If I come across a question that someone else has VTC as unclear without leaving a comment, nothing keeps me from providing that comment myself.