Did you ever find some sort of reasonable 'live streaming' type profile for encoding? I noticed ffmpeg's default seems to be putting more work into this than you'd expect from a realtime encoder which, in theory at least, might lead to fewer collisions than in a 'realistic' situation. I think tialaramex made a similar point somewhere upthread.

I don't think it would make a huge difference but the numbers might be higher/more representative. Another way might be to just capture zoom traffic of someone's Quarterly Synergies presentation or whatnot.

Edit: also thought of a silly upper-bound test for Team Penguin - you get plaintext of every colliding block in the stream. How much penguin can you get out of that?