actually that s not correct, the weight of the chain, cogs, and wheels are overcoming gravity, with loss to friction, in every rotation where as static weight is not being "lifted" or fighting gravity on every rotation but only as much as you climb. Agreed that there would not be a difference in maintaining a speed as long the weight was perfectly centered around an axis, which on a bike it is not, and that during a climb up a mountain that the rider must ensure the exact same wheel rotational speed the entire length up with absolutely no velocity increases or decreases, which is impossible due to our bipedalism. Which would be why I assume many trainers push a good cadence as a priority (reducing velocity changes, accelerating decelerating) to minimize the loss of these damages in every push of the crank. The loss is small but it is still there, in a sport where we speak of grams costing seconds, every little bit matters to some. The damage really comes in in all the minute accelerations and decelerations with every movement of the bike. While the actual mass of any circular rotating object will not change, the relativistic mass does. You can also compare it to engines on cars on a banked track, a conventional piston-rod motor (pistons making rapid accelerations and decelerations within the block), sees a performance loss much greater than a rotary engine where the rotational mass is more centered and constant. If you were to actually look at a graph of the different velocities of a rider going up a hill I believe most would be surprised as to how varying it is.

at least that s what i think