While I haven't yet played with any products using the Bohmer Room EQ, I have played with and calibrated various systems with most other packaged options. I have heard systems and seen the results of using Accurate, but I haven't driven it myself.
The common thread for the best results comes down to being very conscious of the info/measurements you feed into the system, and having the ability to adjust the target while seeing the before measurement the system took. Did you take enough in the right locations? Did you include locations that aren't worth including or problematic to include? You may or may not want to let it correct full range, or you may want to adjust the target curve to largely parallel the response above ~1kHz. By far the most valuable detail is understanding and determining what you might not want the correction system to fix with your specific room and gear.
IMO the crossover function is a separate function to what we regularly call room correction, and sometimes should be included, and sometimes not. Many of the automated systems fail to sufficiently attack out of band resonances without manual intervention. As a speaker designer I'm still of the opinion that any crossover above ~500Hz needs to be developed and specified for the speaker itself in conditions not confused by the room's impacts. The only exception is for systems that might be flush mounted like in a studio or custom installed home system. Even there it's important to have "speaker only" data, and better to simulate the flush mounting elsewhere rather than in room. Subwoofer integration in a 2ch system can be dramatically improved with the appropriate settings and paired with a room correction system.
One of the best results I've achieved integrating subwoofers for 2ch use was actually using a Trinnov Altitude 32 in a dual purpose system with Focal Sopra 2 placed out in the room for 2ch listening where a different seat was optimized specifically for 2ch listening vs an area optimization for home theater use. I have briefly played with Trinnov's Amethyst in a store's demo room, and look forward to an opportunity to play with one in a permanent system, or even using a Trinnov Altitude for a 3-5 way stereo speaker system.
Even with automated systems, the crossover range usually requires some babysitting with external measurements to see how the response of each combines at the listening position. A few of the full FIR filtering systems have a higher likelihood of getting things right on their own if they correct phase/timing to ideally meet the targets, but many look at each speaker or sub in isolation without regard to how they will be blended. This is where having something like
REW & a
UMIK-1 running gives huge advantage to make sure the summed system response actually resembles the target curve you might be editing. I know we want to think that these expensive systems automatically would get all of this correct, but real world observation shows this to be the exception rather than the rule, and most are unaware of this likelihood when they evaluate the first pass result.
An interesting option I would consider, even if just for experiments or getting familiar with such systems, would be the newer
MiniDSP SHD Processor which has Dirac room correction included. While not quite the very low noise floor available from higher end units, at the cost it's a very flexible unit for subwoofer/woofer - speaker crossover and correction.