Difference is that they infer a split based on whole genomes of 3 million years ago, which combined with the present morphological differences, could indicate this species is valid. But as mentioned by others they did not sample in part of the range (Bhutan & N India), so maybe new research would find something of a cline... On the other hand, stable hybrid zones or gene flow after a split are not necessarily indicators that species should be lumped (e.g. brown and polar bears).
And let's not talk about the ridiculous Klipspringer split