I added the second part about the random sample later to the comment when I moved the proposal already out of my short term memory. I hope they use the data from their initial study to test whether the opt-out group actually is different from the group they already get data from.