>Even if the two beams were mirrored into the same trajectory, it's possible they'd refract while traveling due to their different wavelengths and end up as two dots at the end anyway.

Do you mean that they would refract differently passing through an interface, or that the two beams would interfere with each other? It seems like the former should be able to be controlled on the device side as long as you are careful with the optics (though shining the laser through an interface would split the beams, but nothing we can do about that). You could probably cheat your way around inter-beam interference by using a pair of pulse-width modulated lasers set out of phase so that the beams don't overlap and relying on persistence of vision for the laser to be perceived as yellow.