For the preprint --> journal point: if all the bad ones never made it to publication, that would show the same conclusion. Maybe journals are more filtering than enhancement sometimes.
Or, the norm of expecting to publish in a journal means people are motivated to do good work. If you preprint with the expectation to publish you’ll do good work. But there can still be a lot of terrible science out there in preprints. I don’t think what you wrote is evidence we should trust random PDFs on the internet.
For the preprint --> journal point: if all the bad ones never made it to publication, that would show the same conclusion. Maybe journals are more filtering than enhancement sometimes.
Or, the norm of expecting to publish in a journal means people are motivated to do good work. If you preprint with the expectation to publish you’ll do good work. But there can still be a lot of terrible science out there in preprints. I don’t think what you wrote is evidence we should trust random PDFs on the internet.