This week I answered this question, and rather than give a quick answer (technically, just "yes" would have responded to the original question), I explained my reasoning and gave several illustrative examples, so that anyone reading it could see how it applies more universally. I was sure it could count as a helpful answer, as is.
However, a more experienced member has commented that I should have included "supporting evidence", suggesting that otherwise, we open the arena up to mere opinion.
After a brief search, I could find an academic source to back up what I was saying, but was that necessary? I've researched this very question myself in the past (I was an ESL teacher for ten years, and am a native speaker, so I also just know when something sounds right), already knowing that the use of an article like that felt right, but at the time just couldn't explain why.
That past experience allowed me to give the illustrative examples so that it made sense. The explanation itself is what makes the answer credible, rather than the sources I may have used to find it, that I've long forgotten by now.
Even if I know my answer is right, far beyond it counting as opinion, and can explain the logic and give examples, do I truly have to back it up with an academic reference? Or can I at least just say that I'm an experienced ESL teacher and native speaker if credibility beyond the explanation really is necessary?
Providing academic references seems like overkill for something like this, unless I'm making a claim that other native speakers may find dubious. I think more flexibility would be needed compared to making scientific claims for instance, that you'd have to reference some kind of research. As a native speaker, surely if I am very confident that it sounds right, and I can explain why it works a particular way and my explanation makes sense, that's all that's needed?
If this supporting evidence necessity is the policy of ELU, it's unfortunate as it adds an extra step to what would otherwise be a perfectly functional answer. Also, this isn't exactly applied universally. Is it necessary to add supporting evidence for it to be a good answer on ELU, or should we only request it if we actually disagree or doubt the answer or feel like the examples don't hold enough water to stand on their own?
There is another question on the same general theme that's already been asked, but that takes it as a given that all non-high-rep users would give references, without truly justifying it. The question remains - if an answer addresses the question, and does so thoroughly without leaving potential for disagreement, then why would a reference be needed?
Even the answer to that question, which was insightful, alludes to someone like me who was frustrated with the idea of having to provide references, but doesn't really address the problem, only saying that a combination of multiple factors including documentation make an answer on the site "useful".
Forgive me for stating the obvious and pointing out a huge irony, but... that point really could have done with a reference or documentation to back up why references in this context are required in the first place. This isn't a humorous attempt at a recursive loop of references that prove that references are worthwhile, but a genuine frustration at feeling like I'm being told "because we said so".
Fortunately, the answer here addresses my point more directly.