How well can we gather stats on # files, major file types, amount of use, last updates, # users/dowloads, etc. to a given collection? Seems like this might be a useful way to provide an overview when people hit a collection in the ACR.