Wow, indeed that is quite the undertaking. The necessary precision boggles the mind, especially with a 3 color set up.
I wasn't familiar with the Philips Prism, by name at least. This is interesting... http://www.alt-vision.com/color_prisms_tech_data.htm
So I admit that I'm not up on the "latest" in tri-color separation (*cough *cough, when would that have been, 1951? ) but what about reflectors that can selectively reflect colors? Mirrors were coated by means of sputter deposition techniques to produce only reflections of given wavelenghts. This would eliminate the need for filters.
In the book mentioned in my OP this was discussed, but he introduced it more as the "standard" by which all other systems will be judged, saying that this is theoretically possible. So IDK about logistics.
Another question; do the separation filters all possess the same filter factor, and if not, how does one match exposures? ND? Only a consideration with a one-shot obvioulsy.