I think that we need to seperate power drive belt sizing from drive stiffness sizing. Sizing a belt drive to carry a torque load, basically belt tensile strenght times pulley radius along with a "service factor" (dependent if the load is high shock, vibration etc) and you can come up with the proper belt to drive the load that will provide satisfactory service. Yes belt speed is also very important. Sizing for stiffness requires that you know the inertias of the various parts of the drive system and then you set a frequency that you want to design the drive for and go to it. General Electric, back in the 60s, did lots of development on machine tool drives and they came up with a method that allowed you to take all of the inertias of a drive system and "reflect" it back to the drive motor, this inertia along with combining all of the drive components stiffnesses, which can be combined by adding the inverse of the components spring rates allows you to come up with a system natural frequency. So once you were done doing all of this mathmatic "dancing" and you came up with a natural frequency for the drive system.They developed basic rules of thumb as to how much higher this frequency should be in relationship to the drive motor response ability. If you had the drive frequency 3-4 time higher that the response time of the drive motor then the system would be stable. BYW Drive response of a IC engine is pretty slow.
So after all of this "jabber" how does that apply to cam drives, my cut is that a belt cam drive is probably so stiff combined with the inertias of the pulleys and the cam shaft that the belt probably doesn't even see the cam event frequencys as anything that would excite it at its natrual frequency and the belts inherent dampening charateristics are probably a plus. Jackson hit it on the head, strobe lites, load sensors, and testing is what will tell you what is happening.
Rex