More accurate than binaural? - lol no, binaural is how it would sound in real life if its done correctly, id like to see them do it more accurate than that
Did you read the article? Binaural HRTFs are algorithms based on two mics placed inside a dummy head... but nobody's ears and head are shaped just like this generalized dummy's head. That's why we often vary in processing preference and percentage. Supposedly, this should be an HRBF, whatever that stands for, and the algorithms are based on starting with how the brain interprets distance, degrees, and elevation.
Obviously I have no idea how well it'll work in practice, but more developments in this field are a good thing for everybody!