Don't worry, I'm not going to trail off on a discussion about horsepower and torque. I only bring them up to make this point: Even highly experienced practitioners don't always understand the importance of different characteristics.
In our world, we face the same confusion. Look through some recent ads in bicycle magazines and you'll notice two recurring themes: They're getting lighter and/or more aerodynamic. Which is more important? Let's do some comparisons.
All the calculations presented here came from Tom Compton's superb Web site called Analytic Cycling (www.analyticcycling.com). Analytic Cycling performs the computations and also presents the equations.
Cycling Events Near You
It is accepted lore that lighter bikes are better bikes, and that reducing weight on wheels has the best effect of all because wheels rotate. I've heard otherwise level-headed folks claim weight on a wheel's rim counts for six times the weight elsewhere on the bike.
Let's send that familiar bovine to its long rest before doing anything else.
Rotational Weight: Does it Matter?
Weight on a wheel affects the wheel's rotational moment of inertia. The moment of inertia is a measure of how the weight is distributed on the wheel. The common wisdom is rotational weight counts twice that of non-rotational weight. That's because to accelerate a bike, you have to pick up translational momentum, which is what you gain moving down the road, and you also have to pick up angular momentum, which is what you gain spinning a wheel.
The rotating-weight penalty only affects acceleration, and we really don't have to worry about acceleration very much. The effect of a bike's weight during acceleration is small, and the effect of a wheel's weight is very small.
Prove this to yourself: Run up a flight of stairs. Then run up a flight of stairs carrying a wheel. Finally, run up a flight of stairs carrying a spinning wheel. You get the idea: The weight of the wheel, spinning or not, has very little effect compared to what it takes to accelerate your body.
The other and more important reason the effects of rotation matter little is because we don't accelerate much. If you are doing kilos on a velodrome, then worry about it. Maybe. But in the typical 40K bike leg, we accelerate exactly once, with an additional partial acceleration at the turnaround. You can't even measure the effects.