Hi Carl,
First of all, my compliments on your innovative design. In a size and price range dominated by 5" two-ways, you have certainly chosen a road less travelled.
Obviously you have paid a great deal of attention to getting a smooth on-axis frequency response, and the resulting curve attests your success in achieving your goal. I believe John Atkinson uses 1/6 octave smoothing on the curves he runs, which leaves gives very good resolution of small, relatively inconsequential response wiggles. I see virtually no such wiggles in your frequency response curve. What sort of smoothing are you using?
Can you tell me where the crossover is between your midranges and the tweeter? Are the midrange cones crossed over before they go into breakup mode?
I estimate that you probably have no more than 1/2 cubic foot internal volume for the woofer section. Given the calculated efficiency (derived from specified sensitivity and impedance), your frequency response curve shows amazing bass extension for that size box. Is that a true simulated anechoic measurement - no boundary reinforcement?
Your impulse response and phase response curves are something of a mystery to me. If one is correct, I don't see how the other can be. The impulse response shows obvious arrival time differences and phase rotation between drivers, which isn't reflected in the phase response curve. Can you shed any light on this?
Thank you and best of luck with your venture,
Duke