Double or drop the occasional frame I guess? Audio is easier to stretch because it is stored as frequency components already.

Audio is easier to stretch but audio should also never be stretched because we are extremely perceptive to it in a way that we aren’t with our visual processing abilities.

Video should always be kept in sync with audio and not the other way around. You can freely drop or repeat video frames (within the boundaries of reason) without anyone being able to detect it in a blind test, but audio should not be messed with. This is why when syncing audio and video from different sources during production, the master clock is always the audio.

Video conferencing already does massive audio compression (>2x speedup) to compensate for temporarily dropped connections. 30.5s of audio delivered over 30s isn't noticeable if you keep the frequencies the same, which is more than adequate for multiple people watching a movie together.