I think your "THINKS" are accurate, but I still worry about pulling up water from the 20' depths. I know thermoclines will resist water moving from one side to the other, but that confine has got to be weak compared to several hundred gallons per minute of flow in such an open environment. I'd love to see a study that shows the strength of the thermoclines!
Here's an idea to consider (see photo). It is based off of the simple under-gravel filters you see in fish aquariums that are powered by air stones. The overall lengths of the tubes would have to be guessed at, but would not be too critical so long as the inlet is far from the bottom of the BOW. And, the tube diameter, compared to the diffuser diameter would have to be taken into consideration. If using a 12" diameter diffuser, a 20 inch diameter tube would be my guess. The larger the better, but 20 inches would allow for alot of water to be drawn past the diffuser (the diffusser acts like a plug if the tube gets too small an water flow gets choked off). Is this practical for you? I am PRETTY SURE this would greatly reduce the amout of deep water being drawn up to the surface...
EDIT: Something similar could be done using a couple 55 gallon drums with the diffuser centered in two stacked barrels (top barrel without top or bottom, bottom barrel without it's top and holes cut around it's lower perimeter). The J-Tube would be better because it draws it's intake water from above, whereas the barrel rig would draw water from it's lower sides.
