Except the author is not claiming that Latin's masculus and femina/femella are unrelated, but instead that English's male and female are unrelated because "one

Except the author is not claiming that Latin's masculus and femina/femella are unrelated, but instead that English's male and female are unrelated because "one is taken from Latin and one is taken from old French." That the French words are inarguably related and the lineage all leads back to the definitely-related Latin words, would mean the English versions are also related. Given the article's premise of words being unrelated, I'd say that's more of a massive plot hole than a "nit."