You're technically correct (which is the best kind, of course), but for peer-review of scientific papers it's not really relevant. There's nothing _private_ ab

You're technically correct (which is the best kind, of course), but for peer-review of scientific papers it's not really relevant. There's nothing _private_ about the paper, since the author intends to publish it, or about the reviews, since they're going to the author anyway. Only the identity of the reviewers is confidential, and for that email is perfectly sufficient. Well, a standard email client would be far too error-prone; you'd want something like a blind mailing list just so the editor doesn't click the wrong button and accidentally unmask the reviewer.