Sarcs, yes I've seen that study, but there are others as well. As I said, there needs to be more practical research using real-world tasks. Where possible, colour should not be used as a stand-alone indicator (like in the PAPI) because all sorts of things can disrupt the signal, even for colour "normals". I agree wholeheartedly with Arthur on this one, that the T-VASIS was a much better system, as it was not reliant on normal perception of colour, and was less likely to be corrupted by atmospheric conditions/dirty lenses etc. That doesn't mean that colour coding should be removed altogether, as it provides very useful additional information.
The aviation colour environment is far more complex than a tower signal lamp or a PAPI. Really difficult to design a single study, but there are big brains in a lot of the universities that could do it.