Sparks my memory is not as good as it used to be and I have forgot things and I misconstrew things from time to time. Here is how/what I remember.
A multi-viscosity
oil say 10W-40W for instance, means at a cold temp the viscosity behaves like a 10W
oil would at that same cold temp, and that at operating temp the
oil behaves like a 40W oil would at operating temp. This I am fairly certain is a good practical explanation.
To get this characteristic, they start with a 10W oil and partially polymerize the oil, which is attaching strings of molicules together. I believe that a multi viscosity oil actually outperforms a straight weight oil in some load tests because is it easier to squish the single molicules than a chain. I'm not 100% sure this is absolutely accurate, but the general gist is ballparkish, I think.
I would bet more engines have failed to too high a viscosity oil than too low. Viscosity when plotted against temp is very exponential. Typically you see it plotted on a LOG LOG scale graph. Some of the drag racers pushing 35 psi boost into a 4.6 ltr ford and getting 1500 hp at 10+ K rpms were running 0W-5W oil to keep the lifters from pumping up.
Light oils seem offer very good protection. Personally I would not run any more than 10W-40W in anything I have ever owned. I would bet a 10W-30W verses a 15W-50W oil dynoed in the same engine would show significant hp loss with the heavy oil. Every bit of that loss is extra load on the dizzy gear.
At 70F a 50W oil would be darn hard to pump. I wouldn't try it.