This plea to authors to change their behavior ignores why they submit papers to paywall publishers: The prestigious journal's acceptance of their paper helps promote their career.
Academic publishing is not a web host server for pdf files type of problem. Therefore, suggesting authors to upload their pdf to a public Dropbox url, GoogleDrive, Github repo, or their university faculty homepage doesn't solve the real problem. So even if Scihub had a "direct upload pdf" option, that still doesn't solve the underlying problem for getting their paper recognized for good work which spurs citations.
Scihub is a distribution mechanism for pdfs but not a recognition and impact filter for which papers are _important_. This is why scientists keep doing contradictory behaviors: On the one hand, they praise Scihub because it gives them access to papers -- but on the other hand, they keep submitting to paywalled journals to help their career.
Think of game theory incentives instead of hosting pdf files. Journals have the respected editorial staffs to look at their submitted paper and forward it to other peers for review. OpenAccess is a possible option but most OA journals don't have same prestige as the paywalled journals. That may change but it will take a long time.
Q: Do authors actually _have_ the right to republish the final published version from the journal they submitted their work to?
Q: Where is an author supposed to _obtain_ the final, "official" journal-approved PDFs in order to republish them?
Unless I head for sci-hub, I don't have any of mine :(
The big publishers have worked very hard to step this being straightforward :(
> You really don’t have PDFs of your own papers?
Absolutely not. Where would I have _legally_ have got them from? (This was 25 years ago) For my first paper, I have a copy of the paper journal for that month, kept as a souvenir.
> Did you leave academia
Yes
> and delete your data?
No!
That makes sense. Sorry if I was a bit incredulous ;-) I've been publishing for about 10 years, so I'm only talking about the current situation. Things might have been quite different before academic publishing was all online, before arXiv, and before having a "personal website" was a thing. If the publishing agreements back in the day already had a clause allowing the author to privately redistribute their paper, I would assume that if you as the author can get the published PDF anywhere (from a former colleague that still has academic access, directly from the publisher, or potentially even from SciHub), you should be free to then forward it to others or put it on your personal website. IANAL, obviously, so don't quote me on whether it violates copyright to get your own papers off SciHub ;-)