Hi FredK,
I can't comment on the 10-ms statement other than to say that it doesn't make sense, because our ears and brain combine the on-axis sounds arriving directly from the speakers at our ears with the lateral first and secondary reflections, which contribute a great deal to the perceived sound quality of the speakers. I don't see how the 10-ms sounds could "collapse" the soundstage; they'd have the opposite effect, of widening the soundstage, because they're slightly delayed, but our brain integrates them with the direct sounds so the lateral image is widened (the 10-ms reflections are from points beyond the physical boundaries of the speakers, hence we seem to perceive a wider, more expansive soundstage).
The tests at the NRC were all double-blind and the speakers that had smoother and more linear frequency responses off-axis (which were congruent with the on-axis responses) were all judged in the double-blind tests to be more "spacious-sounding".
Regards,
Alan
