Originally Posted by
elcruxio
one needs to keep in mind that these things don't advance linearly. the effort required for a greater average speed on a century requires an exponential effort. So basically going from a five hour century to four hours is much, much greater of an accomplishment than going from a four hour marathon to three hour marathon.
Neither one is linear. Effort required goes exponential in both cases when you get close to the limit of human body. For a marathon, unless you're from Kenya or Ethiopia, the limit is somewhere around 2:10. For a half marathon it's 1:02. Speed is linearly related to power, and we could say that 1:02 half-marathon is something like 6 W/kg, 2:10 marathon is 5.7 W/kg for 2 hours, and 3:00 marathon is 4.1 W/kg for 3 hours.
4.1 W/kg would just about give you a 4 hour century under ideal conditions (tri bike, flat, no wind, no intersections, etc). It's, of course, a bit harder to do 4.1 W/kg for 4 hours instead of 3.
That would also suggest that my first proposal (sub 1 hour 40k TT) was too easy. You can do that by generating 4.1 W/kg on a tri bike just for 1 hour.
But you are right that the return on further effort in cycling is much less visible. Again using my favorite calculator
Bicycle Speed (Velocity) And Power Calculator (I know that it's biased, but biases should mostly cancel out), 4.1 W/kg (good amateur level) at sea level gives you a 40k TT in 1:00, and 6.0 W/kg (talented pro level) gives you the same 40k TT in 0:52. (The actual U.S. record is 0:47:35, set at the elevation of 6200'.)