I gotta say, this topic is completely out of my realm of understanding. From what I got from the article however, it seems that this algorithm is super early in development. I want to see this thing in action with much more complex images such as hair and water.
Anyway, great post that got me interested in a topic I've never explored before.