I'm arguing that end-to-end encryption in a centralized app doesn't actually do better with regards to any threat from its threat model as compared to client-to

I'm arguing that end-to-end encryption in a centralized app doesn't actually do better with regards to any threat from its threat model as compared to client-to-server encryption. It's sort of a fallacy, centralization cancels out any benefits that end-to-end encryption is supposed to bring over client-to-server encryption. So "but they have" or "don't have end-to-end encryption" cannot be an argument.