[jdikatz]

Even if peer review doesn't actually signal paper quality, it's still useful for science journalists to note whether a paper is in its final form or still a work in progress. Referring to unpublished papers as "pre-prints" or "working papers" could be a less biased way to do so.

Also, it's hard to tell whether peer review actually improves paper quality by comparing published / unpublished papers in a world with peer review, where everyone is writing with the knowledge they'll be intensely scrutinized. Without some sort of detailed review process -- even one that's potentially deeply flawed -- researchers would have fewer incentives to be careful.