One question that lingered while I was doing the flow testing was: "how much coolant flow is needed?" This would be a variable dependent upon how much heat (power) the engine is making with an offset for the time it would take to saturate the entire cooling system... but I still don't have an answer that correlates X gpm to Y hp. The variable speed nature of a belt-driven pump does lend itself though to providing the increase in coolant flow as the power also increases. Correct, I did find that the stock VG pump stopped increasing flow above 5000 engine RPM (with stock crank and pump pulleys) and the flow decreased to 32 gpm at 6000 rpm. I did play around with a few different pump impellers (see attached) to try to extend the increase in flow beyond 6000 RPM but all I found were very small changes that just couldn't warrant the expense to produce. I may pick this up again someday when I have more time but for now, I'm tired of water pumps.