This assumes the rules are black and white and there's no concept of bad faith.
A system could very well have rules/guidelines and then have humans review & monitor the system and user-submitted complaints for any abuse, and harshly penalize such abuse with a temporary or permanent ban.
It could end up in a situation where it's technically possible to gain a slight advantage by gaming the system but no participant will risk a complete ban and the system ends up working well for everyone.
I can't remember the precise reason, but if I remember correctly the reason is that because the poster is the originator in most sites including YouTube, while in Google's case it curates links instead without input from the general public. This is also the reason why Wikipedia still only officially operates in the US: the relatively freer publication rules as opposed to UK or Brazil for example shields them from a lot of lawsuits.
There exists no distinction between "publishers" and "platforms" as far as Section 230 protections from online liability goes. It's entirely decided by who produced the content: users are responsible for content that they upload, the sites hosting user-generated content are not. It doesn't matter if the site selectively promotes, or otherwise acts as a "publisher", so long as it's the user's content.
Ah yes, US laws are the only laws applicable to Google.
I should have prefaced that with "outside the United States": notably UK moves the bar to the middle, which will put Google in legal jeopardy (and libel cases!) there.
A system could very well have rules/guidelines and then have humans review & monitor the system and user-submitted complaints for any abuse, and harshly penalize such abuse with a temporary or permanent ban.
It could end up in a situation where it's technically possible to gain a slight advantage by gaming the system but no participant will risk a complete ban and the system ends up working well for everyone.