hi Jack,
the encrypted message is good idea, i believe, & there is a way to identify & block suspect e-mail 'blocks',
but banning certain e-mail accounts altogether most likely will not completely solve the issue, as there are
ways to still use those accounts & make it appear to be that the source is coming from somewhere else -
manual approval is most likely the best way because it is much easier to deal with at that point, rather than
having to go back & delete threads, accounts & block users (more 'pro-active' & less work in the long run, imho)