I was just starting to work on a project to simulate Amplitude division based interference pattern. And I came to a realization, that yes, classic backward ray tracing is incapable of handling wave optics related phenomenon like interference patterns. You can try to keep track of phase differences and everything, but it would turn in a memory hell very quick. But, you know, we don't really need to simulate a wave front or a wave to actually get the simulation right. I came up with this idea of actually doing forward ray tracing to solve it. Because we already know what's going into the system, as the physics experiments are designed to. So if I send the beams I desire, and then calculate the interference pattern by tracing the rays all along the way to the screen. I might get the results. This is because I can just simply trace a ray with this complex phase angle part and at last when they converge, calculate their interference with simple math.

Just so you know, I haven't implemented it right now. But I feel that this might be the way to do it. But this won't work good in a classic scene in Computer graphics where everything is traced backwards, because it's efficient.