Actually, Bill said he used both ROH's for his tests.
But you are right - I forgot to divide t by 2, and ignored the subtraction from R.
For t=.15", the depth difference is about .001405"...
So why is there a noticeable difference in feel? I'm certain I feel it. A lot of other people have said so too. Including tstop4me.
I've noticed other inexplicable factors before. When I measured or calculated rocker profiles and side honing shapes, of different figure skate blades, they are literally close to identical within thousandths of an inch, sometimes less, except for toe picks, within the same brand line. Yet people express extremely strong preferences for one model blade over another. Exactly what can possibly cause a few thousandth's of an inch in rocker profile or side honing to affect skating? I have come up with no physical model that can explain this. I'm reluctant to claim it is all psychological.
Similarly, hollows seem like they shouldn't matter much. No hollow (like a speed skate) means the bottom of the blade meets at 90 degrees to the ice. By your own measurements, hollow only changes this by a small number of degrees. How can that significantly affect how effectively a blade cuts into the ice? I have an unconfirmed theory for that: I think that blades more or less hydroplane across the surface on a thin layer of fluid, much like a tire in heavy rain. But a blade with hollow pushes some of that water out of the way, like a tire with deep tread.
There are physical models, and some static x-ray diffraction measurements, that say the top 40 or so nm of ice at rink temperatures act like a liquid rather than a solid, because the water molecules near the surface are only chemically bound on one side. I theorize that the slight difference in blade angles due to hollow, combined with the weight of the blade, pushes the water layer into the center, much like the tire push rain water into the deep treads. The difference in angle from 90 degrees strongly creates a sideways component to the force from the blade, which is missing without hollow. So, in my model, a speed skate simply hydroplanes, while a blade with hollow pushes some of the water out of the way, into the center, before the top ice molecules can lose cohesion with the rest of the ice, and liquefy as well. Hence a speed skate glides with less resistance, but a figure or hockey blade pushes more strongly. But I have way whatsoever to confirm this theory. One needs a confirmed physical model to explain why such tiny differences matter. Besides, my model falls apart when you edge the blade. A tilted blade should push the water to the side, even without hollow.
(Note, however, that it is fairly common for speed skaters to deliberately create foil edges at the sides of the blade, extending down into the surface, by re-pointing the sharpening burr. Some figure and hockey sharpeners sometimes do this too - but many simply deburr, as do some figure and hockey sharpeners. A foil edge allows a speed skater to push strongly, and accelerate strongly, at the start of the race, while the foil is still intact. Likewise, they edge the blades strongly at the start of a race.)
But, like I said, even if my hydroplane/center-water-push model is right, it doesn't explain anything about how such tiny rocker profile changes or side honing changes can make a difference. Take, for example, parabolic cut blades. There is a huge difference in turning effectiveness with skis, on curved hills, because the parabolic cut can fit the side of the hill, increasing the length of ski that interacts with the surface, and because skis bend to conform to the surface. But most figure and hockey skating is done on flat surfaces, and the blade is not very flexible. So how can .001 or .002" of parabolic cut (fairly typical) make a significant difference? In fact, does it?
Likewise, if the water layer is only 40 nm thick (there is some variation with temperature), it doesn't take much curved length to take the blade out of that water layer, so, yes, rocker curvature should matter. But one should have a physical model of why that matters. Sure, the more curvature, the less length is within that layer. If one does hydroplane, that should push a bulge of water ahead of the point of contact, so the relationship between rocker radius and hydroplaning should be very complex. Should rocker curvature still matter much?
So why does all this matter to Pro-Filer use? Because I believe Pro-Filer creates a slightly different edge shape than typical powered sharpening machines. In particular, hand sharpening speeds give you a longer sharpening burr, which can be reshaped into a longer foil edge, unless you make a point of removing it. This creates a sharper initial edge. So it is very important how you deburr, or polish and repoint the burr into a foil edge, or somewhere in between. It is an interesting question how the suggested maintenance of the Pro-Filer affects the length and shape of the burr...
And how that burr is produced and repointed. If I understand correctly, hardened steel shouldn't be able to deform or bend that much. There have been claims that sharpening burrs must not contain much of the carbon-related crystal structure that hardens steel, because hardened steel breaks if you bend it more than a few degrees, and that the burrs must somehow be closer to iron than fully hardened steel. Huh? How could that happen? Where does the carbon go? Does the effectiveness of the abrasive affect this in some way? To what extent does the same thing happen at the faster machine sharpening speeds?