I think the idea is to have those cranks occur under low-stress conditions. The amount of force exerted on a crankshaft bearing at, say 100 rpm is going to be a small fraction of stresses at 1,500 rpm. So theoretically the cranks occurring at starter motor speed will result in less wear until oil pressure builds than the same number of cranks when the engine is running.What difference does it make if it starts or not? Its going to take the same number of cranks to get the oil circulating whether the engine starts or not. In fact, I would think that after cranking without starting and not letting it build up pressure, the oil has a chance to drain back out of the pump, while waiting to actually start it.
Just crank it up for crying out loud. It'll prime right up, build pressure and be fine!
Does this make a real-world difference: who knows? On the other hand, it's nice to listen to less start-up clatter.