Some of the most interesting aviation research in the past few decades have been around human factors like psychology, perception, and cognition. If there was some substantial effect to having the buttons be arranged in a different pattern, I do legitimately hope it would have been found by now.
Do keep in mind these devices are cost-prohibitive in the extreme to design, build, and certify. The idea of having separate, parallel processes in order to have a different button layout between regional devices creates a thousand headaches of its own, both before and after production. The issue goes even further, in that just the FAA alone requires simulators of these aircraft to have replicated button look and feel criteria that would make your head spin. Is there even going to be a question as to if you're going to have to have two simulators? Will type-ratings be transferrable? Will there be separate differences training and/or currency requirements between the two distinct input methods?
Some or even most of those answers might turn out favorably for manufacturers or operators or pilots. But just having to ask them drives costs up considerably.
These days everyone types on a keyboard. It's way more universal, to the point where a non QWERTY layout has to impose additional cognitive load.
It was common to type on a keyboard also before, it just was a typewriter's keyboard.
Boeing found a way.
