Here are comparative measurements for the Dynaudio Gemini MTM kit. The stock crossover is an elaborate first order design that I have never liked. Although imaging is excellent, the mid-highs have always sounded colored to me. String sound in particular has a moderate "tin can" nasal quality that ruins the speaker for me. I tried biamping them with an active 24 db crossover, which was much better, but left a hole in the middle. Finally, I just experimented with a 2nd order electrical passive design until it measured well. The final slopes are 18 dB acoustic. I think the revised crossover transforms the speaker. The highs are now focused and pure, and it still images well. The measurements show that the stock Gemini does not have very much baffle step compensation, and the mid-highs are a little depressed and rough. The big dip at 7.8 k evident on all these graphs is a flaw in the 260 tweeter that affected early production runs. (But the peak at 10k is a feature of all 260's.) The response with the revised crossover is much flatter, as is particularly clear in the room response curves. Don't read too much into the room response curves below 200 Hz. There are all kinds of location-specific floor cancellation and wall reflection modes.
The crossover point for the revised crossover is about 2.9 kHz, as shown in the reverse null.
Ever wonder what an MTM sounds like when you stand up?