The issue to me isn't whether sources are a "requirement" or not, but rather how we deal with answers in general. Personally, i believe what we should do is:
When you see a dubious statement, call it out and ask the poster if they can back it up. Otherwise, judge an answer on its raison d'être: How well it answers the question.
An answer that is wrong should be downvoted. A correct answer should be upvoted. Everything in between depends on how well an answer has made their case. Now, I agree the existence of authoritative sources or persuasiveness of logical reasoning that supports an answer is a strong factor in its quality. Doubtlessly higher quality answers are better than low quality ones, and I have absolutely no problems with people who are fastidious about providing sources.
Nonetheless, this site exist not as a platform for exchanging research, but to provide answers. Simply demanding (non-Wikipedia) sources, as some members like to do, is unhelpful. Leaving aside the topic of answers based on logical derivations from first principles, sources are only really most useful when they (a) addresses the question at hand, and (b) supports a point in actual dispute.
Unfortunately, what some of our members have been doing is to leave generic comments to "require reputable sources". This is almost totally useless. It doesn't point out what is factually erroneous about an answer. Often it is left on answers that have more glaring issues than lacking substantiation. It is little more than a lazy copy-paste hazing of new users.
Anyone who took the time and effort to write up an answer deserves enough of our respect that critical comments should offer concrete guidance on where and how to improve.
What we can and should do instead, is to encourage sources on factual points of contention.
Exhibit A: Are there any creatures based on breath or air, and specifically stealing these if possible?
This is a poor answer because it failed to separate sleep paralysis creatures from what the question actually asked, creatures that steal breath. At the same time, it correctly points out that the category is night hags. Answers to the question can easily be found by perusing the linked list of night hags.
Yet, rather than the answer's actual failures, the comment instead fixated on sources. This is frankly inane. What is actually in doubt here? The existence of night hags? How is Wikipedia insufficient in proving that? That comment offered no meaningful guidance on improving the answer, when the answer's main failing is that it skirted the actual question.
(On the other hand, I would consider a request for sources on the mouse breath stealing myth appropriate - it seems to be a misinterpretation of a Yichang folklore that attributes sleep paralysis to being tranced by mice.)
Exhibit B: What was the purpose of the Labyrinth built by Daedalus?
This answer offers a concise summary of the relevant myth. At first glance nothing looks controversial. How exactly should a summary of a myth be sourced? Perhaps there are ways of doing so that doesn't defeat the purpose of a summary. It's also possible that the answer got something wrong.
Yet the complaining comment and downvote does neither. It doesn't point out anything factually wrong or in dispute with the answer. It also did not offer any guidance on improving their answer, with sources or otherwise. The unfortunate user who has never ever returned to our site is not informed on how they might improve their answer, while to a passerby, the comment is pure noise.
Exhibit C: Are there any myths we can plausibly trace back to a common Proto-Indo-European ancestor?
This is again not the best answer ever, although it does provide a couple of examples and point out that European and Indian mythology could be traced back a common origin in the Proto-Indo-European peoples. However, while there are several avenues for improving the answer, a comment instead demanded sources for this statement:
By last research, some historians suspect that origin of Indo-European folks come from Urals(Russia), southern area.
Except this has essentially zero relevance to the answer. And while its location is slightly off (but I do believe corresponds to some researchers' views), the mere existence of an idea is hardly in dispute. Sourcing this statement would have been of no material improvement whatsoever to the answer.
Sources are not a panacea to poor answers. We should stop obsessing over them. No amount of new traffic will save this community if we do not build up the content and environment to actually retain users.