Sort:  

Yes it couldnt be impossible. I mean we have Cheetah already finding similar content. A bot could keep a a small database to categorize comments based on amount of used characters. That way spammed comments will stand out and if it detects a certain amount being used more than usual in a certain periode of time it could easily see if those amounts are tied to 1 account and detect spam. If I knew more about programming I would do it myself but unfortunately I dont...