Originally Posted by rm -rf

Yes, I agree with all your points.

The bicyclerollingresistance chart for the GP4000S tests. I added the purple horizontal line. The 28 has the same rolling resistance at 77 psi as the 25 at 100 psi, and the 23 at 106 psi. (And those are the rear tire pressures I've used on these three sizes. Front tires at 15% less pressure.)


Other energy losses
Jan Heine promotes large tires, 38mm, 44mm, etc. Part of his argument is that they save rider energy that goes into jiggling and bouncing the rider+bike. I think it's likely to be significant on rough surfaces. The rider's energy loss doesn't get tested with the bike lab rollers.

I ride smooth tread 38mm sometimes, and they are fantastic on rough chipseal and broken pavement. The handling seems a little more stable / sedate / controlled than the same bike with 28mm. And the tires are heavier, 200 grams more for each tire. But they don't feel slow at the cruising speeds I use when riding these -- any aero penalty is very low, and they aren't slow at all due to rolling resistance on smooth roads.

For reasonable pavement, I'll usually want the 28s. But the 38s are fun, I can ride over just about anything.