If I were allowed to rewrite all the rules at Youtube, I would simplify censorship. Only content that breaks the laws where Youtube is headquartered would be removed by Youtube. No exception. AI would be allowed to "flag" a video for review and a large crowd-sourced team of moderators would vote on whether or not the video is breaking a law. They would have to attest to which laws specifically are being broken. Moderators that false-flag will slowly lose points that count towards a quorum. When a quorum of moderators agree a video is breaking one or more laws, the video would become invisible to the public and handed off to a forensic interface for law enforcement to subscribe to. The forensic queue would summarize who is posting the most illegal material and give law enforcement officials the ability to search for specific content, IP addresses, geographic locations, etc... Law enforcement could un-flag a video if it is in fact legal. Verified law enforcement are also welcome to join the moderator queue and review process. This might be a good option for those having recently left the police force due to current events. If there are laws on the books that the district attorney does not wish to enforce, they could log in and gray them out. A comment will appear showing which district attorney muted those laws.

As for false information, quackery, etc... there would be icons that people could up/downvote to provide public opinion as to something being false information. Post a video of the earth being flat? No problem, go right ahead. People can upvote the "Cool story bro" icon. Quackery icons will slowly de-rank a video from showing up in suggested. There would also be industry specific icons and people verified to be from those industries could double-up-vote icons as well as post comments with a checkmark next to their name linking to their public credentials and contact information. Cost? The verification process would pay for itself. Being a verified subject matter expert provides exposure for the individual, so there is incentive to pay to be manually verified. Being a SME does not grant the ability to remove a video, only the ability to add a comment with a checkmark. Not all SME's will agree with one another. Let the merits of their conversations drive public opinion and emoji voting. Post a cure for covid or cancer? Go for it. Verified doctors and scientists can post checkmark comments and upvote the quackery emojis if they feel it is required. Or they could upvote the green checkmark if they agree. More verified checkmarks equates to higher rank in suggested content.

Accepting that premise is tantamount to admitting you don't have the judgement to be part of the conversation.