An email for a recipient is either ordinary delivered, delivered as Junk or rejected. A recipient can decide whether to communicate to the sender, that the message was just delivered, or delivered as Junk. When a mail is directed to many recipients and the recipients have different opinion on whether the accept the message, the server can at the end of DATA return a message, to which recipients the message was ordinary delivered, to which recipients the message was rejected and to which recipients the mail was delivered as junk. With the latter part, of course, taking the preference of the user to communicate to the sender, that the message was delivered as spam.
The advantage of this approach is, that the preferences of the users are enacted exactly and there are no delays.
A possible disadvantage is, that the rejection message contains, for the senders unusual, different statements about the disposition for each recipient. The senders must be able to understand the language, in which the rejection message was written.
When an email from a mailing list manager is sent to many recipients and the rejection message spells that some recipients have and other have not received the message, the mailing list manager will interpret only the number codes and conclude, that the message was not delivered to any recipient. On repeated non-delivery retries, the mailing list manager may remove all recipients from the mailing list.