Interesting discussion! I've been doing some careful listening to my Sierras and finding the speaker separation fairly critical to achieve proper imaging of a chorus of voices that are spread across the soundstage. I've been listening to the DVD-Audio of Gaucho (superb stereo recording!) and finding that the imaging really snaps in with a slightly smaller ratio of about .65 - just offering this as one more data point. I wonder if this ratio is dependent on the recording . . .