yes here’s a link
Interesting how close the two 'truncated pyramid and parallelepiped’ examples were to the sphere. And better than the hemisphere. Andrew Jones should be stoked.
But note the somewhat limited scope of measurement: 100-4000 Hz and single point microphone. A good start but you’d probably take the investigation further. Interesting also that the discussion focusses on diffraction rather than internal dynamics.
We also have to consider the compromises involved in full-range or co-axial driver implementations. The former especially with respect to suitable SPL and dynamics, the latter with complex diffraction from the drivers.
Look at KEF Blade for example: co-axial treble/mid but separate bass drivers to achieve sufficient dynamics in the lower octaves (Devialet adhere more closely to the sphere but take a similar approach).