I'm surprised it's taken this long for serious scientific work on this matter. Even Darwin, I believe, posited that selection pressures will drive complexity in an arbitrary direction. I.e. that an organ of arbitrary complexity could become more or less complex with time, depending on what suited the pressures. Gould writes about this in his book Full House, which was published in 1996 (and is an amazing read for anyone who loves biology AND baseball!). Progress is a myth propagated by high school textbooks.
I can't stress how much I need to read more Stephen J. Gould. I suppose many theorists suspected that something like complexity by subtraction could occur - but this is the first I've seen someone provide empirical support for it. As I said in the article, the only times I've seen adaptation for less complex structures discussed, it has always been for structures that lost their function over evolutionary time. For me, testing complexity by subtraction could completely change a lot of inherent assumptions about adaptation in the same way exaptation did.