Pre-prepared streaming video will compress much better: amongst other things, modern codecs will send frames out-of-sequence in order to better compress, have motion prediction, stuff like that. Even if this system is willing to spend the compute to do high-level compression, they don't have the data from future frames yet that will allow them to get down as low as you can go.

Equally, because you know so much about the scene, there are many more opportunities to do smart video compression - however, those would be extremely difficult to take advantage of.