Finished the valve between the reservoirs last night, and I have it filled back and running. Still doing a little bit of bleeding, but not enough to impact any flow rate readings.
This picture below shows the valve I put between the reservoirs:
Previously...this was just a connection between the two reservoirs. The idea behind it was to allow more radiator circulation by allowing some of the flow to be diverted from starting the portion of the loop that travels through the blocks...and instead....make another pass through the radiator loop again. Some of the early ideas called for actively controlling the flow through this re-circulation loop with valves and/or additional pumps. It evolved into a completely passive re-circulation loop that was triggered only when the pressure drop for the motherboard subloop was more restrictive than the pressure drop for the radiator subloop. I could impact this with varying the speeds of the pumps in the motherboard subloop vs that of the pumps in the radiator subloop. But generally, the connection between the reservoirs served to send some of the coolant back through a second pass through the radiators...without going through the blocks again.
This re-circulation....could be both desirable and undesirable...depending on your cooling needs at the time. If your blocks are removing an adequate amount of heat from components at low flow rates...the re-circulation loop would be beneficial and serve to reduce your water/air delta's by sending the coolant back through the radiators a second time. However..it does so at the cost of reducing potential flow rate through the waterblocks. So...in situations where additional flow through the waterblocks would help cool components even more....this re-circulation loop was detrimental.
Electrocutor suggested a valve between the two reservoirs so that this could be adjusted manually as deemed necessary....and I was willing to give it a shot.
It definitely was a good idea. If I leave the valve open....everything performs exactly as it did before. No changes at all. With the valve open, and all 4 pumps on high....I get the following:
Motherboard (waterblocks) subloop - 0.9 gpm
Radiator subloop - 1.4 gpm
When I reduced the pumps to low settings....the flow rate drops to the following:
Motherboard (waterblocks) subloop - 0.2 gpm
Radiator subloop - 0.7 gpm
Now...for the results with the valve closed....which is the possibility I have with this change. As you can probably guess...the flow rates for the motherboard subloop and radiator subloop are now exactly the same when the valve is closed...because it is now a serial loop with no subloops. When the pumps are all on high...I get the following:
So the flow rate through the waterblocks is increased (from 0.9 to 1.1), ....at the cost of reduced flow through the radiator subloop (from 1.4 to 1.1). When I turn all the pumps as low as they will go....the flow rates drop to the following:
So....those are some pretty big changes just with adding this valve. My biggest surprise is how much extra flow I'm getting on low pump settings through the motherboard subloop. It shows that when I had no valve (or when its open now) that I was losing a lot of the additional pressure from the pumps in the radiator subloop...by allowing them to send coolant back through the radiator subloop. It doesn't add any pressure to the motherboard subloop. When the computer is going to be idling for a period of time though...it will probably be most beneficial to have the valve open. The additional flow through the motherboard loop doesn't really impact anything at idling speeds....while the increased flow through the radiator subloop should help decrease water/air delta's....which in turn impacts component temperature.
So.....now I have options, and I like it. Thank you Electrocutor for the idea.