What's wrong with a homework-like question? Homework questions are specially written to expand and question one's understanding - it's a totally valid way of learning. If you think someone's trying to cheat, don't answer (or even comment) on the question - the beauty of a website like this is that if people don't respond to crappy questions they won't be popular/seen.
This website is not meant to be all things to everybody. You are accusing this community of being actively detrimental to physics because we have a very strict policy regarding homework and homework-like questions, but that policy is supported and enforced by the community. We have collectively decided that we don't want the site to become inundated with check-my-work style questions, which (as Dale points out) are better suited to a discussion forum. And in fact:
If PSE has a policy against explicit homework help but PhysicsForums has an entire section dedicated to homework help for various levels and subjects, why wouldn't you simply go to PhysicsForums instead? Why should this community change its policies to suit what you personally think they should be, especially if other (both free and paid) sites which cater to those desires already exist in the world?
If it's because you prefer this site's Q&A format, or the general quality of the answers received, or the breadth of knowledge of the community as a whole, consider the possibility that these aspects of PSE are inextricably linked to its policies. Speaking as a sample size $N=1$, I am not an active user on PhysicsForums precisely because I don't like the relentless back-and-forth of a discussion forum. I am looking for interesting questions with thoughtful, authoritative answers, which is why I am here.
I've seen many, many question written-off as one for the maths stack exchange. Yes, the arithmetic/solving is obviously maths - hence the language of physics - but the explanation, the understanding, is obviously physics - don't reject someone for asking how to solve something physical using maths.
I am guilty of answering more than a few questions which, objectively, should have been migrated to MSE. First, migration is not writing a question off. Questions which are purely mathematical in nature are being sent to a community of experts in mathematics which is better equipped to answer the question in a clear and authoritative way. To be sure, there are likely many people at PSE who might be able to answer the question - but there are also people at PSE who can tell you how to change a tire or cook pasta. The point is that purely mathematical questions are not our community's focus, so we send such questions to a community that's a better fit. I can't imagine a way in which this is anything but positive for the question asker.
Secondly, if there is physics involved in the understanding then by definition it is not a pure mathematics question. If anything, this community errs on the side of keeping such questions open (in my experience, at least). Asking how many Killing vector fields must exist on a generic pseudo-Riemannian manifold is a purely mathematical question; asking why Killing vector fields correspond to physical symmetries is not.
Fundamentally, if you spent the same amount of time complaining about someone's question as you did answering it, your impact would be genuinely positive and fantastic. For those who think this site could be some definitive Q&A textbook - it isn't and never will be. Wikipedia will always be wider-reaching and more democratic - and if you want to contribute to a textbook/people's question, go write one.
Clearly this site is not what you are looking for - and that's okay. You are well within your rights to want to change things rather than simply leave, but we are equally well within our rights to keep things the same if the community disagrees with your proposed changes.
Lastly, it is of course true that the community moderation does not always align with my views. Some questions are closed which should, in my opinion, remain open, and vice-versa. Sometimes community members make mistakes or vote hastily, and voting is by no means universally consistent (which is certainly true of me). In such cases, I vote to re-open (with edits if necessary); if the community disagrees, then c'est la vie.
Overall, I think the policies as they are currently are more positive than negative. Disagreements are part of being in a community, and I will accept that sometimes there will be no reconciling my opinion with that of a majority of other users.