Read on Twitter
Why I disagree: for me it's #PeerReview YES! & #Preprint YES! There are definitely not exclusive. A thread 
https://twitter.com/ScheuringLab/status/1304433607423991808
https://twitter.com/ScheuringLab/status/1304433607423991808
There are bad papers that pass peer-review. We have plenty of examples, even on COVID topics. There's not a dichotomy between "useless preprints" and "peer-reviewed article is Truth". We need some journalists to convey that better to the public. https://twitter.com/ScheuringLab/status/1304433608531341314
But the problem already exists without preprints: hype and reporting of excessively broad conclusions, hypotheses portrayed as results, all also occur with peer-reviewed papers as well.
This seems to be a gut feeling, [reference needed] or anecdotal evidence (it did not happen for me). I would counter with more anecdotal evidence, saying I have received several times some good feedback on my preprints. https://twitter.com/ScheuringLab/status/1304433609630248962
Many areas of physics have all that is described here, this seems like a cliché. I'm not getting the argument made… https://twitter.com/ScheuringLab/status/1304433610683023364
This argument (too many papers published) is irrelevant to preprinting: preprints do not mean more papers are published, you just have access to the results faster. Like you would at a conference, listening to talks or looking at posters https://twitter.com/ScheuringLab/status/1304433611781935111
There we agree: open access is not one of the main goals of preprinting. It's just a nice by-product. https://twitter.com/ScheuringLab/status/1304433612851499009
Sadly, full reproducibility and access to data under the open science principles is poor (and very field-dependent). But that's also true of peer-reviewed papers, and there is nothing preventing preprints from being published with accompanying data https://twitter.com/ScheuringLab/status/1304433614181093376
I think the “consensual history of a field” is a unrealistic, idealised view of the process. The literature is full of bad papers, controversial papers, unsubstantiated claims, papers that end u p ignored after a brief moment of claim. https://twitter.com/ScheuringLab/status/1304433615246430208
Preprints are no different, and their citations are merged with the final paper if it gets published. But you acknowledge where the idea you cite comes from, which is by definition ALWAYS BETTER than no citation, or “private communication”.
In the fields where preprints are new, and in a crisis, there are some excesses in communicating about preprints. But the situation will, I believe, improve as journalists and readers get used to it. https://twitter.com/ScheuringLab/status/1304433616328556547
And any time is a good opportunity to restate that a peer-reviewed paper is not "verified and substantiated science" and/or God's truth. In the scientific method, we reach the status of "verified and substantiated science" when it has been independently confirmed and verified!
If you're interested in other positive aspects of preprints (benefits for early-career researchers, partial alternative to conferences, etc.) see my piece in @NatureChemistry ! https://rdcu.be/b4ffZ
