Because people consider what they send in their email to be theirs, not something a faceless corporation has graciously allowed them to utter. People are also beginning to correlate these mega-companies with government-type power, especially over information (hense the use of the word "censor"). Combine these, and you get the outrage.
But IM providers censor malware links and other scams all the time, and your email provider censors or outright blocks spammers for probably up to 50% or more of your mail volume. Clearly people do actually expect certain classes of communication to be actively intercepted.
People expect something that was not intentionally sent by their friends to be actively intercepted. While making the technical distinction may sometimes be tricky, the emotional distinction is fairly straight forward.
You're right, it's an emotional distinction. Let me illustrate. A man operates a gay rights message board that includes private messages. He tolerates alternate viewpoints, but after a particularly bad flamefest he uses a plugin to ban linking to a hate speech site that is especially contrary to the aims of the forum that he pays for out of his own pocket. It extends to PMs. He explains that if you want to trade such links you're welcome to go elsewhere.
But the aim of Microsoft's messenger is not to prevent intellectual property theft.. especially when the users are under the impression that it is to facilitate private communications..
> He tolerates alternate viewpoints, but after a particularly bad flamefest he uses a plugin to ban linking to a hate speech
So he doesn't actually tolerate alternate viewpoints, the problem is in the framing. If the discussion board was supposed to facilitate discussion of opposing viewpoints (like IM network is supposed to facilitate private messaging) then it's not reasonable to ban one side's resources.