An Independent Assessment of Phonetic Distinctive Feature Sets used to Model Pronunciation Variation