When I make a surround mix, it is purely matrix. There is no 4-track discrete version.
Actually, I am working on a way to be able to listen to ATMOS and other discrete systems and hear the side panning correctly without the cogging. If I get it to work, my objection to discrete recordings is gone.
I do not mix, just a listener of provided media...help me understand / educate me.
Strictly from a listener's perspective, by the time the music leaves my speakers, it's a discrete sound....regardless of what it was in the primary media source. It was either a matrixed / encoded source that was decoded and placed into discrete channels or it was a discrete source carried through to discrete channels or it was a stereo source upmixed by a Stereo 2 Surround technology.
I may have a preference for many reasons of the delivery method, but it is fundamentally discrete by the time my ears hear it.
And regardless of the delivery method, it is either going to be a good or poor mix...it may have a wonderful immersive sound field (my definition of a good mix) or it may have isolated sounds from specific channels that fail to create a sound field or just poor use of multiple channels (a poor mix by my definition).
I only have a 4.1 system....so perhaps I am not able to experience something related to your position on discrete vs. matrixed that occurs with a larger speaker array.
But with all things being equal, I do not understand the concept of a matrixed mix resulting in a better listening experience than a discrete version of the same mix.