On one hand I agree that there are far too many frivolous ratings being sent. Ratings based on opinions or because of arguments, because of a clash of personalities, because of differing idea or views, trolling, and so forth, are both inappropriate and counter productive. They cheapen the entire point of the trust system, and serve more as a punishment against the person rather than a useful indicator of trustworthiness.

However, as Loyce has pointed out above, there are plenty of examples of accounts being correctly red tagged without yet having stolen anything or violated any contracts. If your entire ICO is plagiarized, then you are not a legitimate project. If you are advertising impossible ROIs, then you are not a legitimate project. If you are asking for users to enter their seed, enter their private keys, deposit before they are allowed to withdraw, and so on, then you are a scammer. I disagree that we should be letting these users freely peddle their scams when we have the ability to pre-emptively tag them.