Practice makes performance at a given task faster, smoother, more automatic, and more robust. The "practice makes permanent" quote elsewhere in the comments is correct in this respect. Practice does _not_ guarantee that anything useful will be learned–this requires some way to measure progress, some measure of deliberate reflection regarding past failures and their causes (which will allow you to propose new behaviors to practice), and, for intellectual tasks, some deeper analysis of the underlying principles (which will allow you to propose better principles to guide future behavior, even in out-of-distribution cases which you have never practiced for).

The article appears to be attributing the difference to genetics alone, which is neither accurate nor a useful line of inquiry. And the article is attempting to argue against a Gladwellized strawman of Ericcson's work, rather than the actual research, which feels intellectually dishonest. Not a fan of this post.