> I don't think they're [measurements] necessarily all that predictive of how we're going to feel about the sound. <
Well okay, sure. But some things are worse even when someone believes they're better. A perfect example is an equalizer in the hands of a neophyte. We've all known people who goose the bass and treble, or scoop the midrange, or both, and proudly beam as they show off their terrific sounding system. I'm not saying you and Bart are like that! And there are things I might prefer in some cases but still wouldn't want to do.
Sometimes comb filter artifacts can be perceived as pleasant sounding even when the response is made worse. If a comb filter null happens to fall on a boomy or honky or screechy frequency range, the sound could be deemed "better" with the filtering. My guess is the lenses do something more like that than true broadband diffusion which would be useful. Or maybe they add ringing at pleasing frequencies so the overall effect is pleasing.
That's why I asked about measurements. This is the only way to know if the sound truly is made better or if it's purely psychological. I never expected you to provide test data! But what does it tell you when the manufacturer doesn't even have that?
--Ethan