This is just kind of ridiculously fallacious. Surely a slightly inaccurate paper is not comparable to malicious code. That's not a fair comparison to make.
If we were to continue the GitHub analogy, it would be great if ArXiv had a public "issue" list.