I have implemented DiracLive three times, on both sides of my basement, and in my living room. The sound improved in all cases. But, I think optimizing the speaker placement and/or LP (and possibly room treatment?) before using DiracLive will get the best results.
One element that should be considered is bias towards your current (non-corrected) sound. I would live and test the room corrected EQ for a little while before making any conclusions.
The cost of MiniDSP + DiracLive is not large relative to your total system cost or cost of room treatments. IMHO, best bang for buck I could ever spend. True, tactile improvements in sound that could not be achieved with upgrades to components, cables, or recording type. One could argue that DSP also corrects defects in speaker response (frequency and phase). Sorry for sounding like a salesman, but this stuff has really changed my perception of what good sound is and how to achieve it.
Just curious, what elements of sound reproduction can DSP not correct? We often hear that room modes cannot be corrected since they are a function of how the sound interacts with the room after leaving the speakers. I would not be so fast in making that assertion. By modulating the phase of the individual channels separately, should it not be possible to also correct, at least to some degree, constructive and destructive interference? I believe this is what DiracLive is capable of, but to be honest I am not sure since they are a little vague in describing how it works exactly. I would love to hear from someone with more technical knowledge on how DiracLive works.