I have a file number with the German Science Foundation, and an NSF ID. I have an username for the arXiv and a referee ID for PRD. I have a membership number for the APS and the DPG. I have not one, but three social security numbers. Likewise, I have three driver's license numbers. I have account numbers for the electricity company and for the phone company. I have several phone and fax numbers. I have credit card, swift and IBAN numbers, customer numbers, a passport number, a CAA member ID, frequent flyer numbers, health plan contract numbers, and insurance numbers. If you laid all of these numbers end to end, you'd get something close to the US financial deficit or so.
As far as I am concerned, everybody should get an universal ID number at birth. Period. It could be used as a phone number, as a webaddress, as a passport number. Just imagine that: newborns wouldn't get a wristband saying "Hello, my name is Desiree" but "Hey, I'm C3POR2D3X2PI." Cute! You can add that to my ideas how to improve the world.
But more seriously, for some while there's been talk about establishing a unique Researcher ID that would make a lot of this multiplicity in user- and file-numbers unnecessary, and also cure the problem of correctly identifying researcher's contributions if they have a common last name, or if they have changed their name. Chad had a poll yesterday about the recurring problem of correctly identifying an author. Martin Fenner has a good summary post on the Researcher ID, and he is also polling people for their opinions.
I previously commented in my post The Name Game that as much as I might be cursing my name (the full version in my passport is 30 letters long) at least I don't know anybody who shares it, and I usually don't even have to use initials. Stefan however, shares his name, first and last, with a Stefan Scherer from the University of Mainz who is also a Physicist. Not only is Mainz quite close to Frankfurt, the guy is also about the same age and works on an (at least peripherally) related topic. This has occasionally created some confusions.
Thus, as far as I am concerned, having a unique Researcher ID it is an excellent idea and one that I'd like to see implemented rather sooner than later. It might however raise some unexpected questions. For example, who can apply for a researcher ID? Are there any criteria? Is there an age limit? Can I get a number for my unborn children? And more importantly, can I customize my number? Because, you see, I would really like to be a prime-number.