We seem to be getting down to ever smaller nits, which might, to the forum reader, seem to have pretty much nothing to do with the question on the floor, which is are heavy bikes of any use in training.

This nit seems to be does rim weight affect climbing performance, you saying no, me saying yes, this being an add-on to you saying that climbers are faster if they surge their pedal forces, me saying that pedaling round is faster. Which is based on whether or not a heavy bike forces one to pedal round, me saying that it does and is beneficial, you saying that not only does it not, but even if it does that's detrimental. Is this a fair summation?

This paper which you have presented is much better, IMO. In the limited realm in which the theory was tested, it seems to hold up well. However the questions that we are pondering do not seem to be addressed by this paper at all. I think you offered this paper in defense of your contention that wheel rim weight and thus wheel kinetic energy is a very small, almost negligible term in determining the cycling power required to climb. Since this paper is attempting to model steady state high speed performance on level ground and correlate it with experimental results, the term for wheel kinetic energy probably zero, though no calculation is shown. In any case in this use it would be vanishingly small. A meaningful calculation and experimental validation would be to run such a test with different rims and tires on a steep climbing course. In the investigation reported in this paper, all riders used the same wheels and tires, making this even more of a non-investigation of the effect of rim weight on performance.

The Coyle paper you offered earlier had an investigator wondering aloud if it would be possible to design a biofeedback device to encourage cyclists to peak their pedal pressures, rather than pedal efficiently, since this was the takeaway that the investigators were pushing. Such a device exists. It's called rollers. The whole idea of rollers as a training tool is that the inertia of the system must be kept low. Light rollers, light rims, light tires. By doing this, changing pedaling forces during the pedal stroke are made immediately evident to the rider as a change in sound of the roller/tire system.

For a century, roller riders have been told to smooth up their spins by listening to the sound and trying to make it a steady whir, rather than a RRRrrrRRR, varying during the pedal stroke. It's a simple and effective biofeedback device. I've gotten pretty good at that after 15 years of practice. However you are saying that what we've been told for a century is wrong, that we should attempt to peak our pedal force on the downstroke and purposely create that varying sound on the rollers, and not only that, but that heavy wheels don't slow climbers, even if they only hammer the downstroke and specifically try to create in inefficient pedal stroke.

I don't believe it. My personal experience is that this is wrong, the personal experiences and preferences of racers over the past century seem to agree with me, and the findings of these investigators does not support that view, at least so far.

I don't have any experimental data to support my view, either. What one might look at, however, are the films of racers in action. Those who do a pronounced hammer on the downstroke also bob their upper bodies when climbing, due to Newton's Third Law. As we might recall, this was something of an issue in Stage 21 of the 2001 Tour. "Ram" of the 508, the best rider with whom I have ever ridden, has an utterly still upper body. I've tried to copy his technique, though I can't copy his talent.