I was excited to see this paper a little under a month ago, as it confirms some things I've been using as a working assumption for a long time, yet never actually seen any empirical evidence for. The paper is only a preprint (i.e., an article that hasn't made it into an actual journal yet) now...
