It depends on what you mean by a tri- or tetrachromatic display. If your inks or lights are strictly monochromatic and there is only a single display and they match three peaks of sensitivity of the eyes in question, then I do not see how you could.

If otoh, there were more than one display, not all with the same three colours, thugh none with more than three,  or the three test colours had a different spectral  distribution from the visual pigments, or we were only concerned with only a subset of the visual pigments, then something might be worked out.

Which were your assumptions, and what did you have in mind to establish? Knowing that, we might be able to give you more useful answers.

My apologies! I see that I missed part of your question. The RGB screen might complicte matters, so I should expect that one could do something. Exactly what would work best I could not say, but there is room for research. With a full-colour screen with 32-bit pixels something might be done.

You might object that the pixels are only 3-colour anyway, but that does not prove that nothing could be done. None of the visual pigments has a sharp peak of absorption. I guess that something along the lines of tests inviting subjects to arrange blocks of colour as nearly as possible matching their neighbours could do a good deal.