Heck, I'd go further: for a lot of people, all they really want to do is to be able to "peck" notes out well-enough—and have enough understanding of theory—to compose music in some sort of software.

Personally—and this applies to quite a few people I've talked to, as well—I don't give a one whit about my ability to play live, any more than a novelist cares about being able to improvise stories live. I want to build songs on a canvas, laying down one melodic brush-stroke at a time. It's only what happens when I hit "play" that matters.

I've never ran into any type of music-teaching paradigm that has even vaguely addressed my needs as a student.

Even the guy wanting a half decent melody on a midi controller needs to understand to some degree the relations between those notes, both in melodic and harmonic terms. Otherwise the best he can do is hit random notes and hope it sounds good then wondering why it does sound good and being unable to apply that to different situations.

If I never learn how to write, how am I going to produce an essay?

...and then experimenting with short snatches of melody in the attempt to replicate the effect under different conditions, gradually reinventing music theory themselves; and then coming later in life to the study that gives names to the library of intuitions they've built, finally being able to talk about them with people—but happy that they did in fact build the intuition first rather than just carrying around the bulky handles to the concepts without first seeing the use.

Or maybe that's just me.

Its the same for engineering - if you don't read the existing work in the space of the problem you're trying to solve, chances are you'll spend years doing what could've taken weeks.