Yes, the moderators are often vilified instead of praised but this gives a false impression that an alternative Q&A site with different mechanics (i.e. no moderators) would be superior.
The "game theory incentives" problem with your proposal is that the desirable experts who answer the questions benefit from the strict moderation.
Almost every proposed Q&A alternative to SO prioritizes the question askers and underestimates the answerers.
These 2 perspectives contradict each other:
(1) Moderators are useless: The question askers just want to ask their question. There are no "stupid questions". They don't care that their question is a duplicate. Therefore, the mods are just a-holes on a power trip and just get in the way.
(2) Moderators make the site usable: (Unpaid) experts with some spare time to answer questions don't want to wade through duplicate questions or do someone else's homework assignment disguised as a Q&A, etc. The mods aggressively filter out as much junk as possible to keep the expert answerers from abandoning the site.
When discussing "platform economics", economists often like to identify the "constrained supply" because that dictates where to prioritize efforts. In Uber, the constrained supply is the drivers, not the passengers. In dating apps, the constrained supply is women, not men. In a Q&A website, the constrained supply is the expert answerers and not the question askers. (This matches our intuition that generating new questions is "easy" and a larger population can ask them but answering questions is "hard" and therefore it's a smaller population that can do it.)
Therefore, the mods make the site a little better for the people answering the questions.
Another economics thinking framework is Frédéric Bastiat's "seen vs unseen":
- What's seen : a question gets "closed as off-topic". This action is very visible. It feels like an insult to the question asker and adds to the perception that SO is unwelcoming to newbies.
- What's unseen: cleaned up questions make the queue a little more sane for experts willing to spare time looking for (hopefully interesting) unanswered questions to answer. This positive effect on the queue is invisible. It's not easy for experts to comment on moderator actions they don't see because that's the whole point: filter the queue down to something more manageable and make it less annoying for the answerers.
If you deliberately optimize a new Q&A website for the question askers (i.e. no mods) at the expense of the contributors answering the questions, the experts will avoid it. That leads to a hypothetical website with less answers and lower quality answers. That will make it less popular with the question askers which is opposes the goal of making the site better for those seeking good answers to their questions.