Imagine that I took a pic on my cellphone of two different shades of white color samples from a paint store. Even on the worst uncalibrated computer monitor, you would be able to tell that there is a difference, and maybe even say one is slightly lighter, or slightly more blue. But SO WHAT!!
If I were to then ask, "which color is better?", there's NO way you could assess that, because the photo is tainted by the light in the room, the camera used, and finally the monitor you are looking at. The .JPG compression and transmission of the photo does relatively little to make the whole thing meaningless.
It's pointless to blame YouTube's bitrate in any of this. Even if they provided 24/96k, it would still all be meaningless. I would agree that you could say Speakers A are slightly brighter than Speakers B. But SO WHAT!!