Very interesting. I did some google and found this article that explains what you're describing pretty well. I guess now the question is whether the aero benefit of the deeper rim is enough to compensate for the additional springiness of the wheel. Did you notice that you were slower on the climbs?

No not at all. It’s purely a feel thing. When I’m seated, it doesn’t even register. But some climbs around the bay area are truly brutal. One climb near my house is 11+% for over a mile. I end up throwing the bike around a LOT because I’m too slow to stay seated on steep climbs. So I’m probably torquing the wheels more and more frequently than some people.

Nice to see real data. Clearly, just moving to higher wheels won't make you 50% faster. Might make a difference on climbs and such but on the flats, once in motion, a bike will tend to stay in motion!

In a vacuum with no gravity. I've never been able to ride in those conditions.

Yeah, that was a bit of a hyperbole. At least this guys result implies that on a flat ride, a major difference in wheel and bike weight does not impact speed much.

To be fair, I estimate about a 1-1..5kph difference in speed between the Venge/deep section wheels and the steel bike with box section wheels - but the steelie is also a slightly more upright position, and i am pretty sure that contributes a good share to the difference as well.

Just the wheels? Well under 1kph, IMO - you'd be lucky if it is 0,5kph. Even on climbs, a 250gm difference between wheels on a 80kg bike + rider is.. what, a 0.3% weight difference?

That said, all my bikes have carbon wheels on them. They just look bad-ass :)

i guess some folks think they look really cool.. but idunno.. i like the OPEN PRO CDs i had from colorado cyclist the best of any wheel set i have ever used... i dont think the weight savings is that big a deal unless you are fighting for seconds in a stage race... club cyclist dont really earn true bragging rights for winner of the limit sign world championships..or at least they never seem to last that long. but i guess they look cool... to some..

I can see 10W translating to roughly 1 mph at 18.5 mph.

250g on a 300m climb is 75kJ, and at 20km/h equates to 3.75W. I would expect to notice that if you're already sitting at or near FTP. Even if the speedo doesn't show a lot, it could easily be the difference between normal heading up the hill and "man, why is it so hard today"...

Maybe you have a different CdA than i do, because 10W most definitely doesnt get me an additional 1mph.

As for your assertion that you can "feel" an extra 0.3% system weight when climbing at FTP, and that this extra 0.3% weight is the difference between feeling normal and "why is it so hard": if I am riding at FTP, I am riding at FTP and it is going to feel very hard no matter what - the only thing that will change is the speed - by about 0.3% (let's assume a linear W/kg math for now). Maybe that is the sort of difference that you can notice, but i sure as hell cannot.

On a road bike, maybe for a very small cyclist on a very efficient setup. For a typical rider, 10W will buy around half a mph at 18.5mph.

(Also, 10W from wheelset differences alone at 18.5mph would be very much at the extreme end of the spectrum.)

Watts to raise a mass through gravity on Earth's surface is roughly (mass in kg)*(rate of ascent in m/s)*9.8. To get 3.75W from a .25kg load, you'd need to be ascending at about 1.53m/s. If you're ascending at 1.53m/s while riding 20kph (which is 5.55m/s), it means that you're going up a 27.5% gradient.
Doing 20kph on a 27.5% gradient is going to be somewhere in the neighborhood of a 17-18W/kg effort. That's not "FTP" for anybody.

Professional climbers sometimes have FTPs in the vague ballpark of 6W/kg. In realistic situations, that's good enough to buy an ascent rate of up to around half a meter per second. 250g ascending at half a meter per second is about 1.2W.

And for a typical person on a road bike, the power impact of 250g on a threshold climb is well under a watt.