I used an early version of Dirac a few years ago and had some problems with it. Not with its mic or measurements, but something about the kind of speakers I use. I have Linkwitz Orions, which are designed to interact with the room and light it up evenly. Something weird and phase-y happened when I ran Dirac to EQ the speakers. Now, possibly I'm just not accustomed to a smoother response -- I doubt I've ever had a room with decent acoustics since I started out in hi-fi in 1978. But another issue is that in my right ear, I have a hearing loss of several db at about 1-3khz, requiring some balance adjustment to get a stable center image. I couldn't seem to make that work out using the Dirac correction.
Anyhoo, I note that Amir has a big cancellation at about 65hz and peak at 110hz; I have the same basic thing in my room, though not quite as large, with some cancellation at about 50hz and a big peak at 79hz. What I've done instead is EQ through a software plug-in with my Mac Mini server running through Audio Hijack, or else using the new Roon parametric EQ. I just knock down the 79hz peak with a fairly high-Q filter at exactly 79hz. That removes the objectionable boominess. Everything else works pretty well with the dipoles bouncing stuff all over and the speakers well away from walls and fairly close to the listening position. That, by the way, is the Linkwitz MO. Maybe room correction and dipoles like the Linkwitz designs are not sympatico.