You should play with the modeling software at Analyticcycling.com, but there are few if any cases where you can take any 2 real products, one being aero and the other being light, and find that the acceleration benefit of the light one outweighs the aero benefits of the aero one. The problem with light wheels is that your brain can perceive a difference, but when you try and quantify it with data, it is super tiny. The aero gains on the other hand have been processed out by your brain in the first few seconds of your ride...almost nobody feels it, not really.
The other thing to remember is your expectation. We do these really great blind ride tests with different wheels and it is really amazing how terrible humans are at perceiving actual differences when riding. We completed one recently where the rim weights varied by 100grams per rim and the aero difference varied by 20 watts at 30mph, and stiffness varied by a factor of 2 with one wheel at 30N/mm and one at 60N/mm lateral and radial stiffness varied by 2x but doesn't matter. Riders can't see the wheels and have no computer or data on board.. In the forms the riders fill out, pretty much nobody gets it right and almost every wheel was the most favorite of one tester and the least favorite of another. As soon as we began identifying these features, then everybody starts trying to revise the history...it's just human nature to want to be right.
In the end, the aero effect of the rear wheel is reduced 30-50% when compared to the front depending on the frame design. So if a front wheel design saves you 1 watt over some baseline, then the similar rear comparison generally saves you 0.50-0.70watt. Since acceleration effects are so minimal for bicycles, this almost always plays into the favor of the more aero wheel until the weight deltas become ridiculously large. The other thing to remember is the percentage of time spent doing different things. This is one of the hardest things we discuss with protour riders. Often times the key climb or acceleration they are worried about is only a few km long and when the acceleration happens the speeds are reasonably high. 80% of most climbing stages are spent on the flats or descending, and lately we've seen downhill attacks and such that unbelievably favor aero. This decision should be easy for them as ALL the bikes generally weight 6.8kg, so the difference is purely rotating vs static weight and weight position. Aero should pretty much always win, but rarely does. Ultimately is it so hard because the bike with lighter parts seems so much lighter (in their minds), but of course is exactly 6.8kg.
I had a very interesting discussion on this topic with the Schleck brothers after the 2010 TdF where Contador attacked as Andy had a chain problem. The math here was really simple. Had he been on 404's compared to Contador on 202's he almost certainly would have caught Contador on the 22km long decent down to Bagneres du-Luchon, instead, he was only able to maintain the ~40second gap. The mental state of the riders here is that 'I might have been dropped on 'heavier' wheels' but in reality, the bike would weight 6.8kg regardless and since a large percentage of the mass of the rims is at the inner diameter..and the spokes are shorter, the inertial difference is relatively small...small enough to be more or less meaningless even during steep climbing. Ultimately they more or less just didn't believe it...
http://www.SILCA.cc Check out my podcast, inside stories from more than 20 years of product and tech innovation from inside the Pro Peloton and Pro Triathlon worlds!
http://www.marginalgainspodcast.cc