I don't know what else to show other than a simulated piston of the same size vs a measured response. A change in directivity of 1 to 1.5dB to me is quite a lot in relative terms for drivers of this size. As stated above the greatest differences will be seen further off axis (beyond the hificompass measurement angles) and higher in frequency. If the design is striving for the widest and most even directivity these differences are why the BMR is more suitable, the 90 degree off axis angle has the greatest impact on the soundpower due to the weights given in it's calculation.
Yep, the other driver in the comparison is considerably larger, BMR=5cm, TC=9cm. Still the differences are small. The technology of "balanced mode radiator" relies on break-up. Namely a piston mode region below, and a hopefully most chaotic break-up mode beyond a certain frequency. Just the chaotic regime allows for the wider radiation on top. I cannot see how a 5cm driver would break-up as low as 4kHz. Typically the first break up occurs around 10kHz, so way beyond the BMR's operating range in this design. The dispersion feature of the BMR is not likely to be gainful.
You are right to mention the target of the wide dispersion. But as always loudspeaker design is about finding compromises. I doubt that one dB more or less would outweigh the otherwise possible reduction in intermodulation or more clarity respectively.
The Philharmonic BMR has in some respect a quite special architecture. Very wide horizontal dispersion all through, but in the treble, due to the ribbon, an unusual steep and deep narrowing. Don't know if these features are anticipated by all, or put to reasonable use.