I don't disagree but I would say this is half the story. Music theory is a way to describe the patterns that have emerged over centuries of musical developments. As you say, you can't generate "good music" through rote use of these formulas. But the formulas provide a context both to explain what's there and ALSO to create new music within that context. Musical interest comes mostly from using and abusing these rules.
For example, I can predict ahead of time that if I use a V-I progression there is going to be a sense of tension and then release in the music. There is (probably) no scientific reason for this. But scores of other composers have adhered to this principle and so we all understand this concept. And by recognizing this I have a tool with which to control the level of tension in a musical passage.