This question has already been answered here. To summarize what's there, and what I've read elsewhere, there are three options:
Similar to what you suggested, create a custom property, bake your sound to the property, then create a driver with a simple python expression to drive other properties at the required scale. This is what I did. Details in the link above.
Use the Sound Drivers add-on. (I haven't tried it, but it looks nice!)
After baking, un-bake the sound to a normal F-Curve. Then you can scale it or edit details.
P.S. For my project, I tried to do this with the Extended Polynomial Generator F-Curve modifier, but it didn't work... Not sure why, maybe I didn't understand how generators are supposed to work, or maybe they only work on normal F-Curves.