I'm a bit late to comment on Roobarb's statement, but it seems too precise to me.
My grandfather was one of 12 -- six boys and six girls. In terms of of Roobarb's scheme:
First son – named after the father’s father: William, YES
First daughter – named after the mother’s mother: Elizabeth Anne: YES
Second son – named after the mother’s father: James (mother's father, so YES) Hubert (more distant) Thomas (no one relevant)
Second daughter – named after the father’s mother: Esther (maternal aunt) Maude (no one relevant)
Third son – named after the father: Athelstan (no antecedents, but two later examples, including me) Hall (surname of maternal maternal greatgrandmother)
Third daughter – named after the mother: Ethel Helen (neither relevant)
Fourth son – named after the father’s eldest brother: Ambrose, YES, but the uncle he was named after died on the day he was born; more likely; named after his paternal paternal greatgrandfather
Fourth daughter – named after the mother’s eldest sister: Hilda, no one relevant
So, although they correspond to some degree to Roobard's convention it's far from perfect. Incidentally, Athelstan was my great uncle, not my grandfather.
I should add that my knowledge refers more to Devon than to Cornwall, though Ambrose Bowden (mentioned above) was just in Cornwall, Cawsand -- much closer to the border then than it is now.