How does dynos compensate for the increased air flow at higher speeds? The dyno is obviously stationary
This is only relevant for bikes that have effective ram-air systems (not all of them are effective!) and only at road speeds for which it makes a significant difference. At 150 km/h the maximum theoretically possible ram air pressure is 1% of barometric pressure. You can't tune for that repeatably. It's essentially a rounding error. Insignificant.
And in any case ... The Power Commander doesn't have to do anything about it. The bike's stock ECU does something about it and the PC only adds or subtracts a percentage.
The bike's stock ECU knows the road speed of the bike, or infers it from the RPM and what gear it's in, and it knows the barometric pressure and temperature. From that, it can calculate the ram air effect. From that, the calculations inside the stock ECU adjust for it. The measured or calculated road speed and the barometric temperature and pressure are some of the many factors that the stock ECU uses to establish the injector pulse duration. All the Power Commander does is blindly measure what that pulse duration is (let's say it's 1.000 milliseconds, for argument's sake) and multiply it by 1 + whatever the adjustment factor is for the RPM versus throttle position look-up table is ... let's say you have "5" in that table position indicating a 5% desired increase ... then it operates the injector for 1.050 milliseconds in this case. If the calculations in the stock ECU change due to changing conditions and the commanded pulse duration is 1.100 milliseconds then the Power Commander blindly measures that and outputs 1.155 milliseconds (1.05 x 1.100).
The stock ECU does the hard work. The Power Commander blindly applies fudge factors on top of it. The stock ECU will compensate for ram-air. Anything beyond that is a rounding error.
Having said that; running the bike on a dyno in the higher gears where the stock ECU is expecting ram-air and it's actually not getting it can cause issues. It isn't the ram air effect that's causing the issue. It's the LACK of the expected ram air effect on the dyno that causes the issue. A common way to do it is to simply find the air/fuel ratio that delivers best output torque and simply go 1% or 2% rich of that ... you will never notice the difference.
A lot of people think Power Commander maps are only good at the altitude and temperature that they were done at. Nope. The bike's stock ECU - if the programming is any good, which is not always the case! - automatically adjusts for altitude and temperature. The Power Commander blindly applies a fudge factor on top of that correction. A lot of the factors that govern how engines breathe are dependent on the speed of sound (in the intake runners) and this doesn't change with pressure and only weakly changes with "reasonable" changes in temperature. (The speed of sound at +40 C is 7% higher than at 0 C.)
Large positive numbers in a Power Commander map in the 0% column are probably an attempt to solve backfiring, or on/off throttle abruptness. It's not really the right way to go about it ... it's the easy, gas-mileage-killing, wash-oil-off-the-cylinder-walls, fill-the-crankcase-with-fuel, carbon-up-the-rings bloody rich way to do it.
There is only one way to do this right. Instrument the heck out of the engine and run it under various operating conditions and then establish if there is room for improvement that won't cause other side effects. It requires a load-control dyno and exhaust gas analysis equipment. It is a lot of work. Calibrating a fuel injection system properly takes hundreds, if not thousands, of hours. The more short-cuts you take, the less effective the outcome will be. The OEM put thousands of hours into that calibration. Will you?