I like the idea of using larger bearings and shafts to minimize losses, but I'm curious how much of a difference it would actually make.
I could be wrong, but it looks like the rational behind this design is higher efficiency by eliminating any losses from the standard hex shaft approach.
On a standard WCD, there is no space between the wheel and the bearing. So the bending of the shaft is almost nonexistent. Because of this I think your bearing setup will make very negligible difference.
Last season during a meeting in which we discussed what we wanted to accomplish with our off season drive train design, a very respected mentor offered a great piece of advice. That advice was to go after low hanging fruit. This means to go after the small problems that can be easily solved, and as many as you can. I say this because I think you're tackling a very small and possibly nonexistent problem with a solution that demands a lot of resources. I think those resources could be better spent, a lot could be accomplished in the time it takes to design and fabricate your bearing configuration. My suggestion is to identify that low hanging fruit, and to allocate your resources accordingly. In my eyes it's better to tackle several small problems that fix one with a complex solution.
This is just my 2 cents. All that being said, kudos to all involved in this design. You guys identified a problem and came up with a very elegant solution.
Best of luck rumbling in 2015's ultimate assist
-Adrian