Is it appropriate to ask for criticism of an unpublished physics paper one is writing?
Not sure how such a question should be phrased. An example is "What's the worst mistake I've made in this paper I'm planning on sending to a journal?"
I've already asked questions about other's unpublished physics papers and it seemed to go okay. See: Could ball lightening be a form of plasma?
If it would genuinely and honestly be of general interest and utility to others interested in Physics, then perhaps.
I think what we want to avoid here is being "too localized", e.g. a question that would only narrowly benefit your paper and you.
related: https://physics.stackexchange.com/questions/how-to-ask
Personally, I would love to see more questions based on real research (and real publications). In fact, I can't imagine the better ones.
Of course, good one is asking if a proof can be simplified, or if there is a better function to fit certain data, or if one has to take into account additional phenomenon for the correct analysis etc.
Bad ones are only these which are too general or vague (e.g. ''What do you think of my paper?'', when it is not a 4 pages publication with a very novel idea).
Still, I would prefer to receive bad research-level questions than bad (pre) high school ones (ofter already ->well<- answered on wikipedia or another googleable page). As questions with linked draft show that someone put at least some effort.
If your paper tells people something, then it's not a question. If there's something you're not happy with in a paper, then maybe it's not a paper, but looking for a way to make the issue you're not happy with into an interesting question will help even if you don't get interesting and useful answers.
On the other hand, if you genuinely and honestly think it would be of interest and utility to others interested in Physics, then go ahead, ask a question that points to a paper. If others think the question has been asked in an interesting way, they'll give you the five Reps; if they're neutral, they'll leave it alone; if they hate the question or the paper, they'll say so and you'll lose some Reps — or, if it's bad enough, you'll just lose some Reps. If you're making enough Reps elsewhere, answering questions in an interesting way, asking other questions that are interesting, then you can spend some of your credit. There's a danger that you'll become known as the crazy person who has somehow amassed lots of Reps, which likely would hurt your ability to amass Reps in the future, but if you care about that then only ask safe questions, or only ask difficult questions in safe ways. My sense of SE is that there's plenty of pressure to conform, so extra pressure on this particular issue is likely not necessary.
Given the number of papers that go through arXiv every day, there's a danger that SE could be overwhelmed if lots of people took this route, but I doubt that would happen in significant numbers and I think the bridge can be looked at if people cross it. SE would, however, be doing the Physics Community a favor if at least a few papers can be improved or weeded out before they're sent to a journal. Anything that finds disapprobation on SE would very likely not be sent to referees at many journals (though somewhere can be found for most anything).
Ultimately, who cares about meta? — just do it. If you're going to hate it if everyone dumps on you, do it carefully or do something else. If meta says you can do it, that's not going to stop people dumping on a bad question.
Finally, I'd say that "What's the worst mistake I've made in this paper I'm planning on sending to a journal?" is not a great Question. But then, this isn't a great Answer, despite the time I've spent on it (too much, but hopefully I've learned a little from the stops and restarts along the way even if no-one else does).
Best wishes, Carl.
Peter.
