Here are some more calculations.
One thing that has bothered me is trying to understand what throttle setting the ESC is running at for my governor mode setting. So here is something I tried.

1) I set my motor up in my test stand, along with my CC Phoenix ESC (setup for Airplane mode of throttle), an Astro Servo tester to provide the throttle input into the ESC, an Astro Whattmeter to measure the motor current and battery voltage, and a TCM tack to measure the rpm. I probably should just have used my Eagletree data recorder because it isn'
t easy writing down all the info with the prop buzzing along at a high rpm.
2) I assembled a set of 3 battery types: an Evolite 2s1p2500mAHr pack, the Brodak 3s1p 4000mAHr pack, and two of my FMA 4s1p2100 mAHr packs setup in parallel (equivalent to a single 4s2p4200mAHr pack).
3) Grabbed a series of APC Thin Electrics (10-5 through 12-10). These were there just to provide a load at full throttle. I took some care not to overload things, especially with the higher count 4s pack.
Here was the protocol.
For each pack, I first measured the rpm without a prop (no-load rpm). This gives the "no-load" current I_o and the "no-load" rpm. If there were no frictional or magnetic losses, I_o would=0, and the no-load rpm would be just kV*V^battery. Then I put on a series of props, starting at the small size, moving up until I either hit my largest prop or the current was >35A (my ESC limit). All runs were at FULL Throttle, so there is no uncertainty about average and instantaneous values.

As a comment, I mounted the props backwards and ran the motor clockwise. This put the propwash away from all my measuring stuff and made it easier to write things down (I stayed upstream). I had also learned from earlier measurements that simply walking in in the propwash actuall could make +-10Watt changes in the power input.

With these measurments (Amps, Volts, and rpm), I was able to enter the motor into a FreeWare motor analyzer. This was kind of nice because it then calculated some of the standard motor parameters for me. Now I could have stopped here, but I wanted to compile a series of effective throttle settings, so what I did was write a small Mathematica program to plot power-in and power out graphs of each motor vs the rpm, as a function of input voltage. What you see here are four different voltages, from 70% throttle to 100% throttle (each line is a 10% increase in throttle)

To be relevant, the top voltage I chose was the voltage I measure at the end of a flight with my Eagle tree. Also using the Eagletree data recorder, I know what power I am drawing from the battery (Power input) for both level flight (275 watts) and during the most strenuous maneuvers (440 watts) at my flying rpm (~8000rpm). These values I show as a vertical line.

I note that if I believe these calculations, I am flying my Nobler at an effective throttle of ~75% at cruise and up to 80% during the strenuous maneuvers. Notice that there is a lot of throttle still available (it would be the intersection of the 8000 rpm point on the axis with the full throttle line (it is way past 1000 watts---although the battery would have something to say if I really would try to draw that many amps. In other words I think I am way under where I could fly!

Alan,
My 3020-16 has a Kv listed of 812.  Where did you get 880?  Interesting post.

Crist,
When I wound my 3020-16 I also thought I would get a lower kV, but 880 is what I got by actual measurement. "kV" is just related to the voltage you would get if you used the motor as a generator. So a couple of possibilities are:
1) I missed a turn somewhere in the winding--I don't see how I did that, although it is possible that my definition of a turn isn't the same as Scorpion.
2) My "magnetics" are not as good as the manufactured stock motor. By this I mean that perhaps my magnets are a little subpar in strength of I have a slightly larger air gap.
3) A brushless motor is not a "stand-alone" device, it depends a lot on the ESC. So it is possible that my ESC setup (timing etc) is different enough that it influences the kV measurement at this 10% level.

I did test my stock 3020-12 for it's kV (using the my same CC ESC) and that value came out close to the manufacturer's value (I am trying to remember where I put that data!), so there is no reason to doubt as far as I know the stock values.