The smoothest measured in-room frequency response I have ever seen was a SoundLab panel set up in Roger West’s open-floor factory. They ran pink noise through it and I was watching the real-time analyzer display. There was no smoothing and no time gating. The curve was a gently downward-sloping virtually straight line, down maybe 5 dB at 20 kHz relative to 100 Hz, spoiled only by a 2.5 dB up, 2.5 dB down jog at about 500 Hz.
The 500 Hz jog was narrow enough that it might have disappeared almost entirely with 1/3 octave smoothing, which is more representative of what we perceive than is an unsmoothed curve.
I asked Roger why he didn’t use these measurements in his marketing, His reply was, because someone else might get a different curve with their measuring system and accuse him of exaggerating.
When I was a SoundLab dealer, from time to time I'd have other well-respected speakers in the same room. Usually these other speakers were more efficient than the SoundLabs. But invariably I'd find myself turning up the other speakers even louder in an effort to hear the details which were clear on the SoundLabs even at their lower sound pressure level.