It has often been proposed for closing a question to require giving a specific reason (1, 2, 3, 4). As often, the idea was rejected, mainly because it would discourage reviewers from quickly closing low-quality questions (remember that reviewers are not paid for this, nor do they get reputation or any extrinsic benefit).
SO gets thousands of questions every day, and most of them are low quality. As Sturgeon's law says,
90% of everything is crap
By a similar principle, the 80/20 rule (or even 90/10), crap questions come from users new to the SE family of sites, who aren't familiar with what makes a good question, don't have an incentive to read the rules, or include spam ("I'm working on this on my site" etc.). Closing such questions quickly helps keep the site clean.
Users who would be perfectly willing or knowledgeable to improve their questions are also lumped into the same bucket. This is partly because close votes are more likely to be subjective than clear-cut for these users. Experienced users who ask "in good faith" (a moderation concept from Wikipedia) are less likely to post spam or low-quality questions.
...would be to suggest that the reviewer leave a warning comment, rather than an "On hold" vote if the question was asked by someone likely to honestly work on improving it.
How do we determine who such users are? Rep alone would be a questionable, because it might mean that once you get X votes, you get to ask low-quality questions. But there are other metrics: number of accepted edits may be a better one, for instance. The questions is open, but the gist is:
Be more constructive with users likely to improve their question
Whatever metrics we pick, they should means the person has been here long enough to appreciate constructive criticism and perhaps has a good idea about what should and what should not be on the site.