Going back to what Gene was saying. I think Craig answered it correctly. But again, very slight differances in room position can make big differances. Though I really enjoy seeing those scene graphs, I think we need to be very careful in reading to much into those graphs. For the most part, there really isn't much differance between them (I think only one showed a significant differance). I wish there were some way to express the differances as like a % of what we can hear. I would venture to say 10-15% differance in SPL is likely not to be distinguishable, especially at the frequencies we are discussing here. However, I don't know how you would calculate that differance, because volume isn't linear like that and I think it needs to be normilzed to what we can hear.
Still, not trying to put a downer on things. I'm enjoying this more than anything recently. Its just with inroom testing, I place greater value subject impressions. When you take them outdoor, obviously I'll place a lot of weight on the GP data.
BTW, what is your receiver volume when running these scenes? I'm thinking I need to get TrueRTA and see what my twins are giving me at my seat.