I don't think that information transmitted to a third-party is automatically without an expectation of privacy. For example, there's an expectation of privacy in the digits we dial after being connected in a call (PCTDD)- like dialing your account number, routing a call through a calling card company, or routing to a different department/company through the bank's IVR. The government would need to get a warrant to do those searches.

Also, the Supreme Court and other courts have generally protected anonymity on the Internet. That seems to imply an expectation of privacy in e-mail addresses. When we want to post to /. anonymously- isn't that protected? Or, if I'm posting messages related to health, sex, politics, etc, can't I do that anonymously?

Notwithstanding the fact that we're already in a telcomm call on the Internet, I can maybe see how the pen register is analogous to e-mail routed by your ISP, but when it's routed through a third-party like gmail or hotmail, I don't think the government should be able to use the pen register statute to get that information. The same applies for VoIP- it's not the ISP that's actually doing the routing- it's the VoIP service provider. Now, if the government wants to do false friend or pen register analogue at the VoIP or third-party e-mail then fine. I don't think they should be able to be lazy and circumvent the investigatory work simply by saying that someone along the line will use this information for routing. That distinction requires looking into the content of the message at the ISP and then grepping out the potential addressing = a search!