In these pumps, how much of the impellor's output sneaks past the wear ring back into the impellor's eye? A small amount of water sneaking past the wear ring may not be a big deal in itself but does it significantly disturb the incoming water from the suction piece? With that being said, I wonder if the larger impellor gained higher top speed (with the relatively same bowl pressure) due to less water sneaking past the wear ring? Note the water leaving a larger impellor enters the bowl at a deeper point (greater distance from the wear ring).
Maybe it's not sneaking past the wear ring... perhaps the location of the impellor's output in the bowl is the critical factor. To expand on this observation, with the larger impellors, the water has a shorter distance to travel as it is "pushed" by the back side of the impellor's vein to where it enters the stationary vein of the bowl.
I'm probably way off base, what do you all think?