They don't link this data with something tangible like "profits" nor do they really try to quantify productivity in any way, so it is left up to the reader to d

They don't link this data with something tangible like "profits" nor do they really try to quantify productivity in any way, so it is left up to the reader to decide whether the effects they observe actually hurt or help productivity. Case in point, what they consider useful spontaneous interactions between disparate corners of the social graph could in reality be annoying distractions that decrease productivity.