Interestingly, I measure (and tame with a 31 slot graphic) a 12dB peak at 50Hz. I guess it's more than likely that, given what Wayne said about the unreliability of in-room measurements, that I'm creating the problem.

Maybe that's the issue in and of itself - I thought that by measuring at the nominal "sweet spot" I was doing the best thing possible, but there might be some sort of superposition at that location that has me actually damping the overall bass response.

Guess I have to get a bit more crafty....