Doujin works vary a huge amount as to how well-known they are, with some popular enough to get official English translations while others struggle to sell a dozen copies. It would be really difficult to establish strict guidelines for what is on-topic and what is off-topic. Rather than trying to decide, which would be pretty difficult, especially if no one with close-votes has read it before, we should not set any limits in terms of popularity. A question about doujin works still needs to be on-topic and constructive, the same as any other question, but there should not be additional requirements.
However, asking about doujin works which are entirely unknown is unlikely to result in your question being answered. At the moment, we have fewer than 100 "avid users", so if your question is about something that fewer than 1% of anime fans know about, you're better off asking elsewhere. While questions about these sorts of things should be considered on-topic, at least if they meet all the normal criteria, if there's basically no hope of the question getting answered, users can still decide that they think it's a bad question, and downvote it as such.
As for questions about the doujin industry, that's sort of pushing it. People working on doujin works are usually amateurs, and there isn't very much information available, even if you can read Japanese. There are some exceptions (Ryukishi07 and ZUN come to mind), but that is the general trend. As such, while we allow questions about the work itself, questions about the industry are borderline at best. There's a bit of debate if we should even allow questions about the anime industry, and some of these questions have been closed (some have also stayed open and gotten many upvotes, so it's a matter of construction). Questions about the doujin industry are almost universally going to be difficult to definitively answer and very rarely will be of interest to the community here.
Hence, we should not allow such questions in general. There may be rare cases that are allowed in which a question is well-constructed, definitely answerable using information available to the public, and interesting to a fairly large number of users, but those are not the norm.