I'm sympathetic to the idea that a simple question often prompts a simple answer and doesn't require digging through stacks of books or combing through journals. But this is a sub for answers from historians, yes? Even if one doesn't have a formal credential in history (cards on the table, mine is in government) if one wants to answer a question on this sub one should have studied the topic in some fashion. If one can't think of a single even generally related source, just off hand, does one actually possess the background (independent or not) to contribute?
I think this would actually be helpful, because we could see which history books (popular history included) are reaching people. Interested amateurs especially.
We ask that folks provide sources upon request. This is the middle ground that we find most effective. The issue with requiring a source for everything is that it is a. more time consuming for users and mods (having to find a source to attach to every claim, fact, or argument; deleting everything that isn't sourced) and b. wouldn't necessarily result in higher quality answers. Citing Ye Randome History Booke doesn't make the answer better if it wasn't good already.
I do think we need to foster an environment where asking for and discussing sources is accepted, encouraged, and commonplace (not saying that it isn't that way for the most part already.) If you are unsure about somebody's claims or want to know more, ask them! If someone refused to provide sources, report the comment or send us a modmail. We do take sourcing and high quality answers seriously, but there are practical limits to what is reasonable to require of folks.
I'm speaking for myself here and not the sub as a whole.
In many cases there doesn't really exist a single simple source. For example, the history of privacy has three major academic histories--Stones Family Sex and Marriage in England which is a massive 1000+ work with hundreds of citations, the French La Vie du Privee, which is a multi volume work on the topic and covers insane amounts of topics, and finally Keons The Secret History of Domesticity which is a difficult and theoretical book to read even if you are an expert in the field, and also weighs in above 1000+ pages.
Generally speaking, and this is the policy of the sub as a whole, we require sources when they're requested and we try to review those sources. The amount of work that would be required to review all answers for sources would be problematic, however we do encourage people asking for sources.
Extra problematic is how do you define 'simple' source?
I think the current policy of mandating that sources be at hand when required is adequate enough. However, I do agree with shlin28's point about primary sources being used. In the right hands they are indispensable and extremely useful, but a person who doesn't have first-hand experience in their analysis, can hardly hope to be considered a reliable interpreter of them. So in short, I think people who are not proper experts in what they are answering and have not studied the primary sources in depth, should limit themselves to using solid secondary sources.
I agree, though I think my position is in the minority amongst users of this subreddit. Sometimes I think it should perhaps go even further, such as by mandating the use of modern secondary sources as well a discussion of them. Too many times I see primary sources being cited without any analysis, which annoys me a bit since primary sources should not be treated in such a straight-forward way. In an ideal world, I would prefer that every answer contains a dissection of the evidence and some discussion of what recent historians think about this topic.
This can be quite tough I know, especially for fields where older literature is still relevant or authoritative, but perhaps something like a 100- or 150-year rule for secondary sources could be considered as well? I personally try to cite reasonably up-to-date stuff, particularly from the last decade, but I've noticed that this is not often the case for other answers.
That's just me though and I can definitely see why this should not be a blanket rule applicable in all circumstances. Your suggestion is a good one however, especially the part about seeing which books have the biggest 'impact' amongst the general public.
I wouldn't like it only because I'm lazy and my sources are in the booklist for my area, but otherwise I support it because sometimes I see an answer and it makes me uneasy.