On Fri, 11 Jul 2003, Sheldon Brown wrote:
big snip
The length of the hill does get involved with heat dissipation
requirements. A larger rotor will be less prone to overheating on long
slow descents.

Actually, the worst case for brake heating is at some intermediate
descending speed -- at very slow speed the power input (force x velocity)
into the brakes is very low, since V is nearly zero. At very high speed
air drag provides all the braking, the F on the brakes is zero.

David Gordon Wilson did a nice little analysis of brake heating in Human
Power some years ago -- back issues available through http://www.ihpva.org