I remember reading a decade or so ago that early phoneme exposure was one of the most important things to being able to sound native in multiple languages. To me this feels related.
Doing a re-read I'd agree the latter part of my statement is a bit of an exaggeration, but based on how the paper defined the categories, Gc and Ga having the most meaningful impact overall still seems a fair takeaway. Though if you'd like to call my labels for those categories reductive, that's fair.