Shallow crossover slopes? Sure looks it from this graphic.
Maybe. Looking at the vertical polar response graph, a couple of things jump out. The "suckout" at negative polar angle is just barely below the horizontal, which means that the main lobe points up toward the ceiling. Without doubt the two drivers are acoustically out of phase, because the two fronts become increasingly out of phase as the two respective distances (from the listening point to the tweeter vs. the listening point to the woofer) become more nearly equal. (Amir wrote that the reference angle was centered on the tweeter.)
I wish that I had a non-useless understanding of Klippel's Spin-O-Matic, but my guess is that in order for the upper and lower "eyes" in the vertical response to be located equidistant from the reference plane, the reference plane would need to be moved higher, above the tweeter, such that the reference plane will bisect the main lobe, which is evidently aimed up at the ceiling. I suppose you could turn these speakers upside down and then tilt them back a little. But with respect to Klippel, if the main lobe is tilted up or down as appears to be the case with this speaker, I don't understand how the reference point can be selected appropriately without consideration of the tilt of the main lobe, or without consideration of the distance from the speaker. If the main lobe is tilted, then if the lobe is aimed upward, you should need to move the reference point accordingly higher, depending on the distance from the speaker to the mic, shouldn't you? I mean it seems intuitive to me that if the lobe is aimed up toward the ceiling, that if you move the mic further from the speaker, you should have to set the reference point higher compared to where you would place it if the mic were closer to the speaker.
The other thing is that the eye below the horizontal extends over an unusually wide frequency range, from about 2 kHz to about 5 kHz. This of course means that on the horizontal plane hitting the tweeter, the two acoustic fronts are out of phase over this unusually wide range of frequency. Judging from the "Driver components near field", the crossover point is below 2 kHz, which means that the range of frequencies where the two acoustic fronts are essentially out of phase is entirely above the crossover point. I.e., the two wavefronts appear to be in phase at the crossover point but then become out of phase as frequency moves higher. Starting just slightly above the crossover point, the two fronts are out of phase and then remain out of phase for more than a full octave above the crossover point. My understanding is that with some (odd order) crossover topologies the two fronts are out by 90 degrees (in "quadrature"?), however if this is intentional and done intelligently, it ought not lead to a broad-in-frequency suck-out just a hair below the horizontal plane hitting the tweeter.