Sometimes, there can be no answer. If that is the answer, then so be it. Say so. But, in the climate of false information being so prolific on the internet, we need to install the sense of pride that is possible if we insist on sources of information and enforce that rule strongly.
ChatGPT, and similar in development, are examples of how it is even more important.
I have given plenty of other examples in my blog article on checking facts before taking them as gospel. Although, when reading the section on falsities within church sermons, you may go on the search for a better phrase than the one I just used.
The top voted answer in the first link in this question said,
When we should delete or not is touchy; I think people should be given the benefit of the doubt... when there's a reasonable chance their post could be supported by scientific evidence. Leave a citation needed notice and give them a chance to edit, sure.
I may have become a little hard-nosed over sources being required, but for reasons I have laid out, I stand by my approach.
I have never liked the idea of marking unsourced answers with the standard mod notice and leaving it intact ad infinitum. It negates the aim of providing a website that provides a reliable source of answers (where there are answers).
In my view, if a mod adds this marking on an inadequate answer, that question should be monitored. After a couple of weeks, if nothing is done especially by the person who posted the answer, the so-called answer should be deleted.
Maybe the 2 weeks prior warning of deletion could be added to the mod notice? I think 2 weeks maybe too long to a degree, but it gives ample time for the answerer to add enough source citations to save their answer from being deleted.
Downvotes must also be added. The top answer to the 1st link states that
bad content should always be downvoted. Downvotes are good for the site! You should not "stop" when the score is 0; a score of -3 at the least shows an answer is "wrong".
And, as also pointed out in the same paragraph, downvotes are not rude.
While @JeromyAnglim stated in answer to the 3rd linked question
I think that if you have read the literature or have a reasonable understanding of theory, you can often provide a reasonable answer even without citations. Furthermore, in some cases literature is not necessary for a great answer let alone a good answer.
Anyone can spout off an answer without backing claims they have made. I could.
OK, maybe adding sources won't necessarily make an answer more correct, but, it is lazy and, in a sense, it is rude within a scientific stack. Science involves checking the facts. How can the facts be checked if the sources of information are not provided?
For someone who knows the subject matter of the question well, it will be very easy to debunk myths and downright lies. But it will be harder when you are not familiar with the subject. For visitors to this site who have no knowledge of psychology or neuroscience scientific literature, it will be nigh on impossible. It is these visitors we need to protect. With social media, and bad press reporting, lies spread by false information can be further spread to the point where the truth becomes so diluted that it is harder to recognise.
You could say "Look, what is given here is common sense. Why do you need citations for that?". Common knowledge such as the sky being more commonly blue on a sunny day doesn't need citing, but you can take that argument to the extreme.
It was common knowledge that blue mass pills (mercury pills) could aid with mental problems. Even Abraham Lincoln is known to have taken blue-mass pills for treatment of chronic melancholy.
As this is a scientific stack, let's see the research backing bold claims. Even if those bold claims are common knowledge.
It was scientists challenging common knowledge who enabled us to progress and change those false beliefs.
Uncited answers do not belong here.