O/k, didn't knew that was possible. I do not want to infringe intellectual property rights, hence won't copy the actual graphs from Erin's site. But look into the distortion a bit further.
Close to 10% at only 86dB@1m in HD, up to 1% in higher order HD in upper mids / lower treble. I personally deem such to be inacceptable right away. The intermodulation - that's the distortion that generates subharmonic artefacts, hence is not masked by the tones played, rises up to 15% (!!).
This all is an indication of something going out of control entirely. Something to be designed to act linear gets chaotic. You cannot expect to remotely predict what actually is happening. And so, falling back into a subjectivist's attitude for good reason, I would not see such a device as the right tool to convert my recordings to sound. This is all way beyond of what we have experience, and learned to live with.
I have some experience with true ribbons, AMT and magnetostats (the planar design used here). Most of them showed more or less chaotic behavior. Sometimes it is intended see DML speaker, sometimes it leads to the planar to destroy itself. The original maker, BG, was praised to solve the problem to a far extent. But the Bravo is, frankly spoken and taken by the data, a little bit crass.
A final question, what does the Swiss cheese in the title picture mean?