They shouldn't but if you can't have a perfect system then it's best to have a system as close to perfect as possible and that is something we do not have. So ideally an entity wouldn't be punished for requesting a takedown of their own content but because of the insane volume of bogus takedown requests I think it's better to punish those who send them and if anyone accidentally sends a takedown request for themselves then them's the breaks until the system can differentiate between legit, bogus, and self takedown requests.
Furthermore, I'm not sure how YC deleting a post and issuing a DMCA takedown request against content on their own site are the same thing. This isn't about deleting content at all. Whether a site deletes content isn't at issue whether it violates copyright or not. The issue is sending bogus takedown requests to others.
I read parent's post as taking about self-censoring bots like Youtube's Content-ID, and not DMCA sending bots. As far as I know, Google doesn't use the latter.
No, exactly the opposite. I'm saying if you can't have perfect then take the next best thing. In your example the next best thing would be taking an imperfect spam filter over none at all.
Furthermore, I'm not sure how YC deleting a post and issuing a DMCA takedown request against content on their own site are the same thing. This isn't about deleting content at all. Whether a site deletes content isn't at issue whether it violates copyright or not. The issue is sending bogus takedown requests to others.