Not doubting your statements about vertical alignment but would like more confirmation regarding the science behind it. I checked your website and don't see any credentials to back up the engineering designs that you market.
Credentials wise I'm a contributing editor with AudioXpress magazine (and prior to that Speaker Builder), have authored over 40 magazine articles and one book on loudspeaker design and other audio topics, have 40 years of professional experience playing electric bass, own an audio consulting firm that counts among its clients the 6,000 seat Meadowbrook Performing Arts Center, and am a full member of the Audio Engineering Society.
Technically speaking when one is exposed to sound waves emanating from two or more point sources they arrive at the listener from those sources in various degrees of phase, the differential of which being determined by the relative distances between those sources and the listener. Positive phase relationships result in augmented response (peaks) while negative relationships result in diminished response (valleys), the sum total of which results in response that is a series of peaks and valleys referred to as 'comb filtering'.
As an example, two drivers when listened to with a twelve inch difference between their centers and the listener's ears will have a null in response where that twelve inch path differential is equal to 1/2 wavelength (565 Hz) and all the odd multiples thereof, and a response peak where that differential is equal to 1 wavelength (1130 Hz) and all even multiples thereof, with irregular response in between the two extremes. As one moves from side to side in the horizontal plane and the relative distance from multiple point sources to the listener shifts so does the frequency of the response peaks and valleys;the frequency response of the system is therefore different for every person in the audience depending on where they are located.
However, if the radiating sources are vertically aligned then the relative distance from their acoustic centers to the audience on the horizontal plane remains constant, and there is no comb filtering of the system response.
This phenonmena is one that the average audio engineering student learns about two weeks into his first semester.