There is still a lot of links.
In short, the way you are trying to estimate is not considered. And 100 degrees for LEDs - a lot. So they are not degraded faster than incandescent lamps, they must be hot 70 degrees.
Will add some details.
- "In the Datasheet are heated to 100 degrees" is incorrect. Is the maximum, not nominal operating temperature.
- To judge the effectiveness of the radiator only by its temperature - it is impossible, because the thermal transition resistance between led and radiator is an unknown quantity. At high resistance, the led may, conventionally, have to burn out and the radiator will still be barely warm. To judge the effectiveness of cooling in temperature, if only to measure the temperature of the led. This temperature is the only baseline that matters.
- Because the lamp in question is not described, nor is it clear how in General to consider the situation. Because, for example, a short activation, the radiator (a device that performs intensive heat exchange with the environment through convection and radiation) - not needed. I was doing some lighting for macro photography and microscopy for such a regime, where the role of sink carried out the piece of metal with a high heat capacity but low heat (so corny it was cheaper and faster). But for a situation of constant work, this trick wouldn't have worked, because quickly enough the heat sink heated up, and ceased to provide the low voltage of the LEDs because heat dissipation from its surface would be manifestly inadequate.