Heat sinks are rated in degrees c per watt. This means that the amount of heat dissipated is constant and that as the outside air warms, the device, too, warms to maintain the difference.
Not correct. The term degrees C per watt is the thermal resistance a heat sink requires at a given device temperature. A component will generate heat based on its intrinsic thermal conductivity and voltage/current -- which determines its operating temperature. The heat sink design needs to know this value, along with the power dissipated and the expected ambient air temperature. Then the minimum degrees C per watt can be determined for the heat sink required to maintain that temperature as a maximum. As the ambient air temp increases, the device doesn't warm -- it dissipates less heat but still enough to keep the device being cooled at or below its maximum temperature. If the ambient temperature expected is very high, then that requires a different heat sink with a lower C per watt. If the ambient is expected to be very low, then a heat sink with a higher resistance can be specified (cheaper and smaller).
The fins of a heat sink increase the surface area to transfer more heat by conduction. They also are in the path of moving ambient air. The movement of air through the fins removes more heat by convection. The heat sink designer can now factor the shape of the sink as well as the thermal conductivity. But they still transfer heat based on ambient air conditions -- the lower the ambient, the faster the removal of heat. This is the whole idea that the hp people are selling us, minus the ambient air temperature part. It is nothing more than razzle-dazzle heat sink engineering applied to the equipment chassis. My point is it doesn't work that way.