But what-if the 'visual speakers' need to feed-back to the MusicMaker to say "Don't feed me any high-frequency components, because the user is in bright sunlight and their definition is lost, so instead, spread the frequencies below this cutoff level over the full range"?

Don't take that too literally, I ain't an audio-visual engineer, it probably makes no sense at all. But the point is, you cannot predict what the future requirements may be, nor what vertical splits in the overall architecture will makes sense when the future arrives.

It made no sense to use separate subsytems when they made the big valve driven "wireless" set that I remember from my childhood home. And I remember when my father saw his first "separate subsytem stereo", (deck, amp, tuner, eq etc.)--my sister got it for her birthday--he offered to build her a nice all-in-one wooden cabinet to hold it :).

(Does anyone remember the early 80's UK TV ad. for a Sony audio system with John Cleese? The separates came with a stand-alone cabinet with a smoked glass door, and the catch-phrase was something like "Hides all that spagetti hanging out the back!")

Predicting the future is a very inexact science.