Main "problems" are twofold: namely the step in directivity which leads to a bass bloat in the far field - easily rectified by using the onboard shelf (-4dB) and that it has limited output in the lower bass before the limiter kicks in because it's been made flat to 30Hz anechoic - which means even with the shelf filter it will require some manual intervention for most people for it to measure well in a room.
Once you've done that however it is quite a capable loudpeaker.
If/when I have the time I'd like to experiment covering the bottom and or top slot. I have a feeling this could remedy on one hand the sudden transition from constant to omnidirecitonal and aid with the upper midrange off-axis suckout as I'm 90% sure it is a side effect of the damped vents. So you'd trade in some of the cardioid behaviour for a better spinorama - a theory, but I'd have to test that.
I think you would rather have to experiment with the side slots. Look at vertical directivity. The larger distance between midrange and top/bottom slots smoothens the transition to omnidirectional a bit. If they would make the baffle as wide as it is high, directivity would improve, distortion in the lower midrange would be reduced due to less cancellation by the rear slots and increased volume in the housing would make the life of the woofers easier. Additional woofers on top/bottom would become an option, too.
The narrow directivity at 1.5kHz may also be related to beaming of the midrange. At 1.8kHz the tweeter takes over and directivity becomes wider again.
Using the mounting holes on the rear or some specific speaker stand (see Stoneeh review part 2 for inspiration), one could turn the speakers 90° and get the vertical plot for horizontal dispersion. But I don't think it's worth the effort. Maybe as a center speaker for which the Orbits would be perfect.
Looking at the directivity plots of another highly rated speaker, I think the Orbits are just fine. At least they don't loose cardioid control in the lower midrange, have similar variations over frequency and a much better vertical plot.
Now someone really serious about directivity can use the Orbits rotated with a little EQ. And for peace of mind focus on the best looking graph format.
Data may be objective but interpretation is subjective and suffers from to strong visual bias, especially for engineers and others trained at specific visual aspects of certain plots. Same data, different presentation. Now add some positive comment to the right plot and people will react different.
Add a little EQ (same as on the contour plot above) and it looks almost perfect.
Now which plots represent actual acoustic performance? Or is it all just about good looking plots?