I have been looking through the forums trying to find a way to solve the "submit as spam" problem, and while it appears there are many different issues to deal with I thought of a possible solution. The desire is to be able to flag an email as spam after it has passed through SF and send it back for re-processing, however the problem is that extra headers and data added during the forwarding process will create less reliable results.

If the goal is to provide SF with an original copy of the email and a user-verification of spam content, why not just retain a temporary copy of any emails thought to be valid so that if it comes back as user verified spam you can match the user-submitted email with the original and re-process the original as spam. For a small setup like myself, storing all email for some specified amount of time (30 days?, 14 days? 7 days?) is a simple task. Even for larger setups, the software could store a temporary copy of only the relevant information (unique ID, headers, tokens, etc) that could be looked up and processed as need.

Just my two cents, as I have grown weary of copy/pasting spam emails and updating keyword lists. This seems reasonable enough that I could even create an external database to store the information as it leaves SF, match user verified spam emails to the original source and then resubmit the spam email to SF using a keyword or IP blacklist.

Thanks, and keep up the great work.