The W3C process does not strictly require independent implementations. It's a good safeguard against issues, and not having it opens the spec up to questions, but this has been covered elsewhere in the case of WebRTC:
Multiple independent protocol implementations exist (and the protocol part was handled at IETF, through their processes), so interoperability on protocol level has been shown even if browsers happened to pick the same one.
The W3C spec describes the wrapping JS APIs, sufficiency of that spec has been shown by the browsers independently building those layers around a WebRTC library. Especially given the working real-world use WebRTC already has it those two layers combined are a good argument that the spec is suitable.
Multiple independent protocol implementations exist (and the protocol part was handled at IETF, through their processes), so interoperability on protocol level has been shown even if browsers happened to pick the same one.
The W3C spec describes the wrapping JS APIs, sufficiency of that spec has been shown by the browsers independently building those layers around a WebRTC library. Especially given the working real-world use WebRTC already has it those two layers combined are a good argument that the spec is suitable.