Will,
I am certainly not an expert where you have gone, but I think there are some things to do.

In the Phoenix ESC, I think you want to play with the timing. I "think" what that does is start and hold current longer as the motor rotates. Typically as the rotor magnetic field lines up with the permanent magnet magnetic field, the torque drops (it equals zero when they are aligned parallel--lets call that 0 degrees). So it makes sense, if you are trying to conserve power, to cut the current a little early, and hold off a little later as the rotor passes the 0 degree alignment point.

However in your case you want to even squeeze some torque out near that point, and the way to do that is to have the ESC hang in there a little longer. The current will also increase naturally at this point, because the back emf (which fights the battery voltage) also drops to zero at 0 degrees (that's another reason for the "frugal" ESC to cut out early too).

So you will draw more current, and get a little more torque which should spin the prop faster. How much I am of course not sure.

I guess that the original sensored setup could be adjusted to increase/decrease timing too. I think there is another effect in that when the motor commutes, there is a natural impediment (actually due the the ratio of the motor inductance divided by the winding resistance) to getting the current to flow. So increasing the timing in that sense means turning on the current as early as possible in order to have it actually full on during the high torque part of the commutation cycle. With the old brushed motors, I think that was the only adjustment (like advancing the timing on a spark engine to get the fuel air mixture burning as soon as possible in order for a complete burn in a high revving engine. Since you are also tuning high rpm's that may also be the main effect in going to  "high" timing.

As I understand the Castle Creation ESC timing, the ESC already "senses" the nominal timing for your motor. I guess it knows how fast the current is ramping up in a commutation cycle by actually measuring it -> but I'm not sure tho'). But if you put in the highest timing, that will tell the ESC to "go for it"!.

I don't know if you will get anything with adjusting the PWM frequency (the throttle control frequency)--since if it is at WOT anyway, there is no switching for a throttle going on --as far as I know.

Finally I think the lipo manufacturers are coming out with pretty low internal resistance lipos--the 35C and 45C rated ones. That might really be the best thing to also look into. Another thing to do is to monitor the high performance forum over on RCGroups
http://www.rcgroups.com/high-performance-9/ , and see what they are doing these days. I am guessing you may already be doing that.

Hope this helped out some.

Hi Alan,

More testing.  The setup is Aveox 1415/2Y- six A123 cells fully charged-Castle Phoenix 80 ECU with Hublin timer.  Phoenix shows solid red LED so I'm getting full throttle.  Prop is a thin blade 8/8-  Accurate very low resistance Ampmeter, and digital voltage readout at battery terminals.

Instead of 80+ Amps, 19.17 Volts and 20+ kRPM as determined by electric motor calculator, I record 65 amps, 14.5 Volts and about 15 kRPM.

My Questions:  Could the low RPM and Amps result from too low an advance selected in the ECU?  Could the low voltage reading be an artifact of back EMF and ripple in the circuit and be incorrect?  If so, would a big capacitor across the voltage sensing leads improve accuracy?

I was concerned that a botched in-line soldering job might have locally increased resistance in the battery (possible a rotated end cap), but the battery is cool in this area and all over, as is the motor and ECU after a 40 second run.

I tested the battery into a heater load and saw almost 20 Volts into 80 Amps.

Any suggestions??

Will Stewart