Comment by miemo

It's reasonable to suspect that the party in charge of handling the audio wouldn't have considered that latency to the end user between the two streams would make their release not synchronous.

They may even have thought that a few seconds could still be called synchronous, unaware of the advantage this presented in this particular case.