True, didn't really think about that one much but the delay really does seem small. Must be under 3ms.

.... I just measured it and the delay is 0.8ms - That's with both the graphic and parametric EQ active in dual mono mode using analogue inputs and analogue aux outputs.

The MB2 Xovers from woofer to midrange at 380Hz. So at 380Hz this would (I think, had a few beers!) produce a phase difference of 109.5 degrees. Thus if you add an additional delay of 1.83ms to the DEQ it will be in phase again at the crossover point.

Of course the crossover isn't just at an exact frequency but I would think this is accurate enough to try it and see what happens to the frequency response, which would be a good indication of whats going on with the phase higher up the crossover slope.