Cynical_Stranger A trained model would be all you need to handle this. Python handles this beautifully with its array of machine learning libraries.
Train an AI model based on the absolute deluge of content these spammers are giving, apply it to new discussions as they are created, and if they are flagged by the system as "spam", notify the appropriate moderators for immediate review, after which it would be taken down based on its content.
No need for reporting anymore, No one gets hassled, and the moderators that should be doing their job, well, they would be, and it wouldn't even be much skin off their back for doing so either.