The reason we did it this way is because we did not have stands. At home, I have my M22s directly on a credenza-style entertainment center, so I had no stands to bring. Also, since the tweeter/treble carries the most localization information, we felt stacking them the way we did was the best way to do an even comparison. Sure, it's not a scientifically perfect comparison. As Ken posted in a follow-up, he discovered that the M50's position in the room was less than ideal during our comparison.

We were aware of the difference in vertical position, so we pretty much ignored the height aspect. Our conclusion that the M22 create a better soundstage was based on the width of the soundstage we experienced, not the height.