Chomsky's theories are irrelevant to that finding.
Specifically, we know that infants are hardwired to look for co-occurring cues of all kinds.
They're working around the clock to sift through all their simultaneous sensory streams for patterns, and this also includes trying to induce how to move their body based on random movements.
That's General Intelligence, which is what Chomsky tried to claim we didn't use for language.
He was wrong. The so-called Language Acquisition Device is pseudoscience, unmotivated, and not in evidence now or then.