Been there, done that!
In that case, a 'pinpoint bearing' and a 'straight axle and bearing' would have much the same friction for the same load. That isn't correct.
You've postulated a very light unsprung bogie vs a heavily sprung bogie. I did make the point that the force on the sprung bogie could be equal to the weight of the unsprung bogie. If you can make a 10 gram unsprung bogie work then we could make a 5 gram bogie with 5 grams of springing and 5 grams added to the smokebox work or even a zero gram bogie with 10 grams of springing and 10 grams added to the smokebox. :-)
If at the extreme you have 300 grams on the leading drivers and zero grams on the rear drivers, you will lose no tractive effort, but you have reduced your current collection to one axle and you have a loco that will not reliably track in reverse. Current collection and smooth tracking also count.
Whatever weight you claim will work for your unsprung bogie will also work for a sprung bogie. In fact, the sprung bogie should work with considerably less downward force.
Stiction in model bearings is greater than rolling resistance. From my experiments, stiction in pinpoint bearings is about 150% rolling resistance.
I don't have the space available for prototypical curves or train lengths. My locos pull trains in proportion to those the most powerful loco can handle at scale speeds on my layout. As there are sharp curves in the hidden section and a steep gradient on the open section, this requires that locos be able to pull decent loads.
Regards, Greg.P.