So, the people complaining at an all hands are likely to be completely ignorant about the facts on the ground. They're not working on these systems or processes. They're just gullible enough to believe what they read on the internet, and take action on it.
The team responsible for the account suspensions / content takedowns / etc on the other hand will have the numbers to show that they're doing a good job, and the expertise to predict what the implications of doing the policy changes asked by the complainers would be. Not just how much it'd cost directly, but the second order effects on abuse from making different tradeoffs.
So let's say that you want to give high-touch customer support to billions of free accounts? Not only is it going to be fabulously expensive, but it's going to be abusable as hell. The abusers will quickly learn just what kind of sob story will have the best chance of fooling the humans, and get their accounts falsely unbanned or even use it to hijack the accounts of others. The only way to avoid this is to make sure the customer service reps have no agency. But then you're paying tens or hundreds of millions / year just to have humans execute a script.
And these predictions will be a lot more credible than those from random employees parrotting social media posts and making vague claims about potential brand damage or loss of trust. Ignoring the complaints is going to be a pretty easy choice for an exec.