Ball / dhnguyen:
Well after reading most of this thread and all the mods and improvements I have a couple of comments that may explain why some mods get better results than others.
There are two that you may have already noticed to boost performance besides using a double prop:
a) The more inlet area the more flow.
For review sake is to say that the design issue being addressed here is that for the same amount of volume the inlets area shall roughly be equal to the outlet area multiplied by the ratio of the increase in fluid velocity.
In other words, if you increase the water velocity by a factor of 10 then the inlet openings sectional area shall be 10 times larger than the outlet area!
Taking Ball's idea of the reducing fitting as the housing I will just turn it around and use the larger diameter on the pump side while the smaller diameter towards the outlet. This configuration will allow for more and longer slots for the inlet than can be possible by using a stright housing (a coupling or a piece of PVC pipe) and of course more than using the reducing housing with the small diameter toward the pump.
b) I think it was shnguyen ho noticed that when the slots or inlet orifices extend forward beyond the location of the prop the flow is increased. This could be explained by the creation of an eductor effect. Slots at the sides and slightly forward of the prop will allow for the higher flow in the outlet and not the prop itself to dragg in additional flow. So by playing with the lenght of the slots you can increase the flow further but at some point it might reduce the flow velocity so the trick will be to find the optimal lenght for the distance to be reached by the outflow. Longer aquariums you may want higher velocity but less flow, shorter aquariums you may want higher flow but less velocity.
There are some other areas that could be experimented with, as an example: The flow coming out of the prop follows the pattern of a corkscrew swirling around. This swirling effectively shortens the reach of the flow. That cen be partially counteracted if the front brace supporting the end of the shaft can be made as two or three vane shaped flat surfaces installed parallel to the housing. These flat vane surfaces will serve three purposes, will support the front bearing and also will reduce the swirling motion of the discharged flow and work as stoppers in case the start up is in the wrong direction.
So there it is, those were the coments I wanted to make in case you feel like experimenting a bit more.
Enjoy!
