Why don't the major email services have a way to separate email they know is spam from email they think is spam? If they did this then false positives would be less troublesome, and they could even tweak their filters to capture a few things the currently slip through. I know that this is possible and some people already do this sort of thing, but why don't the big three (Google, Microsoft*, and Yahoo) do this?
Imagine if Google had a "possible spam" tag where items that scored within a certain range in their spam filters would go. Then you could tweak whether you wanted these items to go to your inbox or your spam box, and they would be easy to filter through in either one. Or, maybe if Yahoo showed the percentage of which they were convinced that the email was spam. The default view in the spam folder could be sorted by this field, ascending, so that the least spammy emails would float to the top.
Since no spam filter is perfect, you have to choose between more spam in the inbox or more false positives in the spam folder. I'd rather have some control over the threshold for this, it'd make me more confident that I don't have to sort through the stuff that I know is spam just to get to verify that nothing of value got lost in its midst.
* I haven't used Microsoft's offering since shortly after they bought Hotmail. To my knowledge this post applies to them as well, though.