I downloaded the free Trial of Dirac and played with it for several hours. My wife and I compared the results Dirac provided subjectively and objectively with a separate measurement setup. One of us was at the remote rack location toggling the filters on and off in the software in real time while the other listened. Here is my personal experience:
Dirac isn't for me in the end for the entry price. I A/B tested with my wife, with several different filters, each less intrusive than the next. The best attribute we were able to achieve was tight, dry bass. But if you have heard really, really tight, dry bass, well, it sounds weird and unnatural. MINIMAL processing was my preferred setup for all filters. Toggling the filters on and off quickly showed obvious, although not always better changes. I was only EQing most of the range by a few DB -and it was too much. I could never get every region sounding "just right."
My experience with the Dirac Calibration software was great. It is very intuitive and quick to generate and save filters. The companion Dirac Audio Processor Controller on PC is less than bug free. I have new found respect for those who run an HTPC or a computer audiophile rig. What a pain in the nutz it is to get everything to play well together. Had a few "just walk away" moments for sure....
All in all this was a great experience. I learned about how my speakers react to EQing. Extending the high end treble range causes obvious problems. Flattening out the midrange seemed to shrink the feeling of envelopment in the space. Tightening up the subwoofer response yielded pretty good results, but there was now a pronounced "Hi, I'm a subwoofer" sound to every bassline. The mains and subs were no longer seamless. Asking my speakers to do more or differently than they were designed to do with Dirac made for a lesser experience, but better measuring system. The one thing Dirac did seem to improve was the dynamic impact of transients after impulse correction. But the difference was akin to toggling the volume up a hair with the remote with no filter applied.
I think the tried and true passive acoustic treatment methods are less intrusive and natural sounding than my experience with Dirac -but in a bare room the results may have been very different. I could have probably spent more time with the filters, but it seemed there was no way to EQ small portions of the frequency range and leave the rest alone. The sliders only allow a low\high pass filter arrangement. I would like to try another PEQ that lets the user focus on key ranges and leave the rest alone. I have determined my room has comparatively minor issues.
They measure much worse than they sound. Chasing the remaining issues down with Dirac software was two things for me.
1. Enlightening
2. Soul suckingly un-fun
I learned a valuable lesson. Boiling results down to raw numbers and graphs with Dirac boiled me down to a frustrated person. So thankful for the free software trial. Moving on!

Here is how my main speakers and subs averaged over the 9 measurement positions before Dirac was applied (blue) and the post calibration predicted outcome after the default target filter was applied (green.) I was able to verify the in room response closely matched the target curve at the listening position post default calibration. In the end 9 filters were tested.