Huh...I think they are confusing correlation with causation. It’s very likely students who are good at math, take more music classes.
I agree - correlation makes more sense.
I do remember tho’, in 5th or 6th grade, we did an experiment in our class. We took a series of mathematical tests, some with music (classical) playing, some with no music. The tests were not identical, but all on the same difficulty level. The scores on the music playing tests were higher across the board than the non-music tests. Not a huge difference, but definitely notable.