Agreed, this is important, if only for confirmation value. Before this, we didn't have hard evidence that the large scale collection was grabbing content of no

Agreed, this is important, if only for confirmation value. Before this, we didn't have hard evidence that the large scale collection was grabbing content of non-suspects in addition to the to/from data. I'm not sure what rationalization is left to hide under, but I bet that people will still try to say that this isn't so bad.