The only case where the current E2EE security model is changed is when the protocol gains a feature to enable the client to attach a cryptographic proof about the message's content. That would violate the non-repudiation properties of the Signal protocol, but it would make reports more believable as you can't fake them any more with a modified client. As I've said I don't know the implementation, but it's entirely possible that they didn't include any cryptographic proof and keep it equivalent to screenshotting.
Signal achieves non-repudiation by allowing forgery of messages after they have been transferred and verified by the receiver. It is not possible to undo that. You would have to make a special exception at the time the message was transferred which wouldn't be practical in this case.
This all assumes that you believe that establishing forgeability actually achieves non-repudiation in the first place. Facebook might want to pass along the signature information anyway.
If there is a risk that the user doesn't know that the correspondence is sent if they hit the report button, then fb should add a warning, but the feature itself is fine IMO.
This feature does not change the privacy story for WhatsApp in any way. You already trust users you are chatting with to not copy or screenshot private messages and share them. You already trust the app to not forward or store your messages without your consent.
[1] See bullet point "Your Messages" under section "Information We Collect" on https://www.whatsapp.com/legal/privacy-policy.
~~Even worse, 3 letter agencies~~
Forget I said anything