Variable Cam Timing....

[jfaubion]

>OK.  So I'm not a genius.  However, it might be possible that when 
>the valve train is released may be important -- i.e. that there are 
>times within a cycle where the tendency of the cam is to "run away" 
>from the crank.

This might work on some engines. It would probably be very difficult to 
judge the timing of such an event at high RPM.

>  Some sort of centrifugal arrangement may also work with a spring 
>that pulls the cam back when the engine decelerates.  

This could be a possibility. 

>I would think that motors and gears would be far too complex to 
>actually work reliably.

Yes I agree the KISS principle is a better plan.

John