I got back to Greg yesterday.
Two of my comments got combined somehow, so I was probably talking too fast.
1) The reversed labels for power that I was talking about were on the Gyro sensor not on the camera. See
http://www2.usfirst.org/2006comp/oth..._Operation.pdf for details
2) There is a mislabeled diagram of the CMUCAM2 board that has the pan/tilt servo jumpers reversed. That only affects the direction of the servos if they are plugged directly into the camera board for use with Labview, not if they are plugged into the RC for use with Kevin's tracking code.
For the proper camera power hookup look at pdf page 17 of Kevin's CMUCam2 Workbook
http://kevin.org/frc/CMUcam2_workbook.pdf