My head hurting idea is to actually increase the diameter of the hub from the leading end to the trailing end of the impeller. Water does not compress. So in the design I am attempting, the volume of water entering, has to move faster to get by the restriction created by the increasing diameter of the hub.

This is similar to what the "compression" cone at the back of the drive does to the water stream.
Just a little more agressive. It is working for a gentleman in Japan, Yoichi Takahashi.
So if can figure out how to duplicate his work, I can imagine great success.