If one measures the room, the speaker placement, and the listener position… then one can ray trace the bounces to get the delays from the reflections relative to the direct path.
If one then does a correlation the combing will show the reflections.
Usually one wants to take out the early ones, so the sides off the walls, maybe using a mirror, and ones behind the speakers and maybe behind the listener would be the earliest.
I would do the sides and then behind the speakers and then the behind the listener… and then from there who knows.
This is a great situation in which to use the correlation (or autocorrelation) to put numbers to things.