Get well soon. 
What RGB colour space do you use? As we all know, it depends on the space. Trouble is we have to traverse many spaces starting from the camera’s decided by the manufacturer, each time abstracting a bit more. There is preliminary talk about Camera spectral sensitivity data but these are gathered data from other researchers. Until we can understand, verify and generate them ourselves, they aren’t very useful for our current workflows.
Watching the video, I am reminded of the paper. Seems to be an empirically derived method that is “suitable for” ACES AP0. Isn’t perfect but better than before according to the authors. So not quite black magic, but in the just works department, which I bet annoys you exceedingly.
I think it is a good start. Thinking through the problem is more important than finding a solution at this point because then you understand more of what is going on: the black box becomes a grey box, and hopefully becomes a box that is fully transparent, robust and well understood. My math went down the drain soon after university, so I can’t help you develop your mathematical ideas. 
All the best. Sometimes rest is all you need to make new discoveries. 