Dennis,
You can fake out the motor programs by simply using 70% of your nominal battery voltage to get an idea about what is going on. That's what I do with DriveCalc.
Onto your question:
The easiest thing to do is just follow the glow tradition and ask others what they use. But here I will try to say a few things.
1) Motor kV and battery selection go hand in hand with rpm selection. Choosing one sets up the other. So that's relatively straight forward. I take nominal 3.7V/Cell, * number of cells*0.8*kV should equal where your rpm selection. You can use 0.7 intstead of 0.8 to be a little more conservative with headroom.
2) Motor size---Lets look at a AXI 2826 and a AXI 2820. They are basically identical designs, except the 2826 has a 26 mm long stator and 26 mm long magnets compared to the 20 mm ones of the 2820. If you wind each one with the same number of identical gauge copper wire, then the kV os the 2826 will roughly be 20/26 times lower than the 2820 motor. So to get the same kV, we would wind the longer motor with fewer number of copper turns. Since there are fewer turns, we can then increase the wire thickness (lower gauge) to fill up the same space.
What this means is that the 2826 motor now has the same kV as the 2820 with fewer winds and thicker wire---so its winding resistance is lower. At the same current to make the torque to turn the prop, the larger motor will make less heat. Sounds like a win-win situation!
Go for big!
But we left out the detail about magnetic losses
. Every time we commute the current in the motor, we are changing the field inside the stator. This itself creates heat (hysteresis and eddy current losses). The more iron in the motor the more the losses (at the same rpm). So here the bigger motor begins to lose. Also we are carrying around more weight to lift in the vertical.
If you look at the motor calculations, you will see that for any constant voltage input, there is a sweet spot--where the output/input power peaks (just efficiency). It is a relatively broad peak, and it seems to occur at about 80% or so of the max (no-load) rpm point. Above this rpm and magnetic losses dominate, and below it resistive losses dominate. I'd argue that you would like to sit at this sweet spot during level flight.
The efficiency at this sweet spot is on the order of 80% or so maybe higher or lower. So what this means it that if you are inputting 300 watts of power, 60 watts are going into heating up your motor. The main way to shed that heat is through the outer shell. Notice that a longer motor will have a bigger surface area to shed heat than the smaller motor (and also will most likely be a bit more efficient) and so will heat up less.
It is this heating that sets the manufacturer's rating of most motors. At some point, the motor can't shed the heat and things will begin to go downhill fast. Most of the magnets used in our motors will begin the lose (permanently) their magnetism as they heat up. If this occurs, the kV will begin to actually go up. kV rising allows more current to flow into the motor, raising the heating more......bad news time.
I think most of us in CL don't see this problem unless we severely undersize the motors. I use a Scorpion 3020 (similar to AXI 2820) in both my Vector and Nobler. I have used a 3014 in the Nobler, and it felt fine, temperature wise after a flight. It was a bit less efficient (harder on battery) but was also a less weight. I could probably use a 3008 in the Nobler and get away with it, although I would be more worried about efficiency than burning it up. If I flew bigger planes (like Strega), I am sure the 3020 would be on the small size, and 3026 would be a better choice. I am not sure I would need to go even longer to a 3032 or to the next diameter size up, a 40xx motor.
I would claim that for outrunners, most motors in the same size range like a 28xx or 30xx (these are stator diameter/lengths) are basically electrically equivalent in their non-stressed performance. And that is where I think we want to use them.
Sorry for running on so long!