That's not far from how it actually works. 90% of what I read is from arxiv, and before it has been accepted into a journal or conference. Most of the time that I'm visiting a conference website is to submit a paper or get the proper citation for the paper I'm writing.
Like I'm not sure what you expect, just dumb people to read arxiv? If that were the case then no one would be submitting works to arxiv. The reason works go there is because researchers are reading them. They then talk about the works to their peers, post on twitter, blog, or whatever. And guess what, that's also exactly how works that are published in journals and conferences get passed around because there's a million of them too and no way for people to sort them. No one is sitting around ranking papers. That process happens by citations.
Are you aware that the origin of Arxiv was as a pre-print archive? People read stuff on there because it was freely accessible drafts of published work, and in my field that is still the most common practice.
Yes, but I think you're being too strict with the definition of draft. It is not "draft" as in "first draft of my paper" it is "draft" as in "we submitted this to journal/conference and are just making widely available" or "here's the same copy that is behind a paywall, fuck jstor."
There's a reason most papers that end up in journals or conferences don't get a revision, or if they do only get minor ones. In general it is the same exact work which got sent to a journal or conference. So that is either the final draft or near. I do not want people to confuse this with "early" or "first" draft.
The history of arxiv is a few physicists getting annoyed that, like many of their peers, that sending per-prints (papers awaiting review by a journal) by email was burdensome. This was about just making access easier. It was about literally making peer review easier. Because let's be real, your colleagues reading your work closely is better review by peers than a few random people who are annoyed about having to review and know it is easy to just dismiss and reject.
Like I'm not sure what you expect, just dumb people to read arxiv? If that were the case then no one would be submitting works to arxiv. The reason works go there is because researchers are reading them. They then talk about the works to their peers, post on twitter, blog, or whatever. And guess what, that's also exactly how works that are published in journals and conferences get passed around because there's a million of them too and no way for people to sort them. No one is sitting around ranking papers. That process happens by citations.
Who do you think the end user is? Laymen?