There is a conceptual distinction: Encryption in transit vs. encryption at rest. You may send the packets encrypted to the server, but if they are not encrypted on the server's file system, anyone can read them.

The real question is, why do you think governments make such a big fuss about citizens having access to military grade encryption?

There have been audits of e2ee implementations, and the algorithms used also have some objective properties. I don't think that I have ever heard in cryptography discussions that backdoors are so widespread that the discussion is moot. I have only heard, time and time again, the opposite.

Even Apple, in this very occasion, opted to ditch the service rather than backdoor it, and in fact takes the UK to court over this. I think that the opinion that this is all for show is a tad wild, and not very well supported in this occasion.

Like every cryptology book starts with the adage "There is cryptography that prevents your little sister from reading your mail, and cryptography that prevents the government from reading your mail, and we will talk about the latter."

https://blog.cryptographyengineering.com/2025/02/23/three-questions-about-apple-encryption-and-the-u-k/

On the other hand, not all implementations are created equal. Telegram was recently under fire, and there is a lot of variance in e2ee implementations in XMPP clients, IIRC.