Should be a given (Score 2, Insightful)

This should be a given. If you try to block spam, you are going to block some legitimate messages. Hopefully, your ratio of blocking spam messages against legitimate messages is good, but it will never be perfect. This is due partly because spam itself is subjective. A lot of spam messages can be picked out and determined to be a spam message by 10 out of every 10 people. But for some messages, its not that simple. It's just real subjective. Then you're asking an algorithm to use subjective logic to determine whether a message is spam or not and problems just occur. Like I said, for the most part these filters work pretty good, but its not going to be perfect and anyone that thinks so, is just not thinking straight.

I am not opposed to some degree of flagging an alleged spam message, but to discard it without the end user knowing about it is where issues begin to arise. By flagging a message, the end user is able to use their own discretion to determine whether a message is a spam message and they can do whatever they want with those messages.

This isn't to say that RBLs and spamlists are a bad idea, just if you implement one of these, then be prepared for some type of backlash. Perhaps in some cases an RBL is necessary, but to think that using an RBL you are going to stop all spam and all of your clients are going to be happy, that's just wrong.