But this actually matters! While your example makes it sound indeed silly, full "life-cycle assessments" and counting externalities matters. Permit me to take your example to an even more farcical end: imagine there was a correlation between bikers eating beef and bus commuter being vegetarians - in such a silly world the bus commuters would probably be much greener even though the bus is a gas guzzler. The problem is that with nuclear these analyses are really complicated, and should not be dismissed out of hand. I do believe that nuclear energy is very important in the fight against global warming, but such dismissals do not help convince people of it.