I believe they were claiming that children previously unexposed to the concept of voiced/unvoiced //hard/soft consonants would understand the latter terminology without explanation

Uh-huh, yeah, sure. The way I've heard hard-soft used has to do with whether a grapheme {c} / {g} is pronounced as a stop /k/, /g/ or an (af)fricate /s/, /dZ/. Should be an easy experiment to set up. All we need is some children who haven't been exposed to English grammatical pedagogy.