I spent some time researching and testing the Ninja Pan speed (or lack of). My intent was to write some code for a PIC to see if I could drive the stepper motors at a faster rate than X10 uses. I took apart a Ninja and pulled the motors and the RF reciever out of it. I used a PIC development board that I bought from Micro-Engineering labs a while back. I wired a prototype board with similar drive circuitry for the steppers that X10 used on the Ninja. I wrote some code to exercise the X10 gearmotors such that I could change the pulse widths on the steps to see what they were capable of. Sure enough, they performed just like the manufacturers website said they would...
The bummer is that with the reduction gear ratio in the motor itself, coupled with the final drive ratio in the Ninja housing, the step rate is already at it's maximum. They chose a motor with a really high gear ratio and a low maximum step rate. Unfortunately, there is no margin left to increase the step rate in software. If it is increased even by 10%, the motor starts missing steps. Bummer!
From here, the only way to increase the pan speed would be to find a more capable motor with the same form factor (which is a long shot) , or make some gears with a different drive ratio for the final drive in the ninja housing. Both options would exceed the worth of the Ninja. The motor manufacturer does offer the motor with different gearing ratios if you want to buy 25,000 pieces or so...
On a different subject, I used the RF receiver from the ninja to decipher the command set from the pan and tilt remote if anyone is interested. I wrote some code for my PIC development board that will decipher the remote control signals and process them.