You'd need to combite the spectral response of each film layer with the spectra of the LED. I'm not at all convinced that narrow band emitters will record at all well...
An led which emits only light in one of the films spectral bands (ie affects only one layer) cannot be recorded accuratly - the film only records the intensity FOR THE LAYER.
There's no way to differentiate between two different LEDS that trigger the same band to the same extent. If a green led only triggers the green layer, then there's no way to know WHICH green it was, just that it's green and how bright. It'll be green on film, but not the same green. Any LED that's completly within the green band would look exactly the same on film.
Typically we get away with it as real world sources have broad spectra, and trigger all three layers to some degree.