userfriendly,
I get your drift, and they are to a large part self compensating. Higher viscosity means more local heat generation, and the average viscosity in the bearing is lower than it would otherwise have been...all things being equal.
But they aren't equal...when designing the things, you have at least three iterations between shaft power loss, flow rate, and oil heating to get in the ballpark, and a few more when you get close to find where the design converges.
Typically, and the classical design curves are all Newtonian, in spite of the additional heat generation within the bearing due to viscosity, the viscosity and MOFT will be higher.
More temperature rise across the bearing surfaces (including piston faces and the like) mean that there's more temperature to drive the heat out of the sump walls.
Not sure if the mental gymnastics make sense, but of you have a look at a couple of design curves (pick the L/D - 1/4, it's close), and the first is the oil film thickness, under constant load. Min F isminimum friction, Max W is maximum load...
r = shaft radiua
c = clearance
N = Radian per second (RPM equivalent)
P = applied load, force on projected area
u = viscosity
Make all of them constant, bar u, and see how the parameters slide along the 1/4 line.
versus the relationship on the side leakage