But it's easier for them to extrapolate data and say, hey, we couldn't get emails of xxx thousand users, so we lost xxx amount, vs our argument that they probably lost a number of users who were annoyed by the popup.

Except, you can count conversions, but you can't "measure annoyance"