Assuming the center of the room already matches, in principle one point is enough to establish a roto-translation and another one to also establish a scaling factor if needed. I have to think about it… Also I don’t really use room scale, so I cannot test it for you. There was a similar question some time ago which was somehow related:
https://forums.unrealengine.com/development-discussion/vr-ar-development/1362635-4-17-vr-rotation
Look at the solution there, maybe it helps you.
Anyway, what happens now? Why do you need such a complicated system? Isn’t the regular calibration/centering enough? Or am I missing something?