Yes, I've read that and disputing if the things mentioned there are really such blockers. I was specifically referring to the point 1.3.2 in my suggestion to use "envelope recipient domain" for verification as I really do not see it as impractical as the RFC supposes