We'll need to handle resolution changes in MediaRecorder. I'm for now thinking of a basic case where we are just recording frames in their native resolution. We might want do some spec work to also allow for scaling the frames before they're encoded. There are two things that must be done for this: 1. When we detect a resolution change, reconfigure the encoder. 2. When we are recording a track containing a video format that has to be converted to YUV420 before being fed to the encoder, we'll have to resize the conversion output buffer on resolution changes (at least when it becomes higher than its current capacity).

Another (hacky?) alternative is to render to and capture from a canvas. This is the most flexible option although perhaps not the simplest/most convenient and I don't know if it is the most performant either.

Mass change P2->P3 to align with new Mozilla triage process.