It's not perfect. Preprints are even less perfect. Just having a big finding based off of a single peer reviewed article is also a very bad idea. You really can

It's not perfect. Preprints are even less perfect. Just having a big finding based off of a single peer reviewed article is also a very bad idea. You really can't get much certainty until you've got dozens of papers and meta-analysis finding consistent threads, but people don't seem inclined to wait for that before reporting unfortunately