As a reviewer, it's really painful to get a paper that's completely wrong (especially if it is over claiming) and that already has a large download count on arx

As a reviewer, it's really painful to get a paper that's completely wrong (especially if it is over claiming) and that already has a large download count on arxiv. I'm all for putting camera ready copies on it, but submitted drafts really shouldn't be there.