I threw one wheel on my trifilar pendulum, and used the formulas described on the above mentioned wheelfanatyk blog post to calculate final results.
I used the same wheel, my ENVE Classic 45 laced with ACE3 spokes to a Paul High Flange front hub, using ENVE rim tape. I used the same tube with each different tire, and none of the tires were inflated to any pressure, they were just installed onto the rim.
I did three runs with each tire option and averaged the results.
I show results with no tire and tube (that'd be just the wheel with rim tape), a 23mm Open Corsa EVO CX, a 25mm wire bead Schwalbe Lugano, a 23mm Continental Gatorskin, and a 37mm Michelin Protek.
So, here's some background information/ comments/critiques on the "experiment" :
The end result, the watts column, is watts of energy required to spin that wheel up to 30km/hr in 10seconds. These numbers do not account for friction in any way - not bearing friction, not friction between the tire and the ground. These are just looking at overcoming inertia to spin the wheel in a theoretical situation in which all we are doing is spinning the wheel. It also doesn't really account for aerodynamic drag in any real way - given this test is done, in my case, in a garage and not a vacuum, I suppose there is some wind resistance, happening, but we're probably not achieving speeds of 30km/hr so that would obviously ramp up more at speed.
This also is obviously only the amount of energy required to spin that wheel by itself up to speed. The wheel, in the real world, is part of the whole bicycle rider system, so % variance between these wheels is lessened when taking the whole system into account, but it's worth noting the actual watts variance is not changed.
Results show that every time you go to spin that wheel up to 30km/hr in 10seconds the wheel by itself requires 3.1 watts of energy. The 23mm Corsa tire (and tube) require an additional 2.6watts of energy of the wheel by itself; the Lugano requires 3.4 watts additional; the Gatorskin requires 4 watts additional; the Protek requires an additional 10.3 watts to spin up to speed.
These are HUGE % variance between each other when just looking at the wheels by themselves. They're a considerably smaller % variance when taking into account the whole bike/rider system...but perhaps not an insignificant amount of actual watts. How many times in the course of your ride to you accelerate and decelerate? Saving two watts per wheel (so 4 watts total) every time I go to speed up is worth it to me to run Corsas over Gatorskins....and I'd wager it's something you and I can feel.
You can also check
here for a spreadsheet of mine I'm updating with wheels I test to give data. It's not currently a very good spreadsheet and I haven't yet figured out a formatting that I like to show it in a way that makes clear obvious sense.
Also attached is a a photo of a Michelin Protek mounted to an Enve wheel...
in case anyone wanted to see that sacrilege