Based on my experience, I would guess that in most typical rectangular shape room like my two rooms, 6-7 dB variation (peak to peak) is easily obtainable it the Editor app is used with Ratbuddyssey or some other paid 3rd UI to reduce the dips and boost minor peaks slightly. The process should take less than an hour, depending on how much trial and error needed, and the time to run REW in between, to see the real efffects.
To achieve 2 dB or less peak to peak, i.e. +/- 1 dB in the 20 to even 150 Hz range, I had to spend many hours over days, not worth it and there is no way I could claim any audible benefits, vs +/- 3 dB peak variation in that range, especially if I put a Harman tilt in the deep bass range.
That's why I believe all those claims about how great RC, whether it's Audyssey, Dirac Live, ARCG or even Trinnov are mainly (say >75%?) based on people's eyes and brains, and/or their measurements such as REW graphs, if the graphs look great, it must sound great!!), not their ears, though I do believe for those who apply RC/EQ full range to 20,000 Hz, audible difference would be real, but in those cases, the users better know what they are doing... On this part (about the >75% brain based, not ears), I bet even @AcuDefTechGuy
might agree with me.