I'm gonna chime in a little late on this thread. I read through all of your thoughtful comments and a couple of observations from our side of the trenches:
1. We use room eq fairly regularly... let's face it, a crossover is a series of equaliziers. We often take up three or four bands of eq to 'create' a proper cross over slope for our systems, so technically a factory tested setup on Griffin monitors alreayd is "eq'd"...
2. We agree whole heartedly that equalizing for comb filtering caused by early reflections is a loosing cause..
3. Larger boundary and modal issues that cover a reasonably large range (1-2 feet) around the mix positions are candidates for cut only eq.... You can absolutely lower modal ringing as it is percieved at the mix position by lowering the output from the system (and yes, changing the linearity of the system) with some cut eq// Remember the "System" is comprised of the wire, amplifiers, crossover, loudspeakers AND the room!!
Sometimes there are valid and practical fixes that can be made to improve a control room's response. Sometimes there are not! this is just a fact of space available, budget, practical construction variables and client resistance...
4. I am NOT a fan of room equalization, but given the fact that modern DSP controller units like the BSS Soundweb London, DBX Driverack and Lake processors all offer crossover functions, micro delay adjustments, multiple bands of eq and limting, we feel that if we can tweek the system "in place" to make it more linear as a COMPLETE system (and the mix position) with reasonable restraints, then we will do it. The limits of how much eq and how it is applied are all based on experience and common sense with understanding of what is CAUSING some of the non-linearity in the final response. Remember, it is the acoustic response that the engineer hears that he bases his mix decisions on!!
so, I guess I fall in the middle hear. Ideally I'd prefer to do NO room eq, but the nature of (our) crossover designs includes so much "eq" to make a proper acoustic crossover response, that it's already in there... Lars can speak to this more elegantly that I, but he and I typically agree on what mix position anomolies are candidates for eq and what ones are not.
As far as the Auto Adjusting DSP systems from JBL , Genelec and others, we have some limited experience with them. What I can tell you is that the last Genelec 8050D surround system we setup in a room of our design, sounded more 'musical' than any I've setup recently AFTER the autocal... so, subjectively I have to say I like what it did. The autocal does very little above 2K (I think) and it mostly 'corrects' lower octave issues. You can also make multiple measurements at multiple locations and 'average' them if you feel this is valid...
"accurate respone anywhere"??? no way... not possible IMHO...
I just believe there is room for this technology to work hand in hand with proper trapping, reflection control and room design. There is ABSOLUTELY no substitute for proper room proportions, speaker placement and lots of broadband trapping!! that's my soapbox and has been for years!! I'll be talking about this next week at the TOC (I'm covering easy DIY basetrap designs for the masses)... and we believe you can never have too much of it!!
Anyway, there it is, I've chimed in and probably disagreed with many (on the eq/DSP issue) but this is our real world experience along with the theoretical considerations weighed in there.... I don't know, but I think there is room for the technology...