Hey Matabele,
I develop MyCdA and would like to help out.
1) airspeed - can you send me your .CSV file so I can try and see why it won't import.
2) as far as how airspeed is used in calculation, in the virtual elevation equation - it's assumed that v_air is equal to v. When air speed (v_air) is available it's plugged into the CdA part of the equation replacing ground speed (v). It's pretty much how Golden Cheetah does it having looked at their source code as well.
3) can you send me the aeropod .FIT file - assuming it contains all the air speed measurements and stuff. This way I can add support for Aeropod files in MyCdA without needing .CSV
I've gone further and played with weather data from online services such as DarkSky and OpenWeather. DarkSky can give you minute by minute historical data which is pretty nice. Having this data and combined with GPS coordinates of the activity I can calculate heading and combined with wind direction and speed reported by the weather station I do some math stuff to calculate air speed and yaw. In some situations I've seen it work nicely - ran my own Ironman AZ file through the test - where wind direction and speed changed throughout the ride.
This assumes of course that you're not drafting, the weather data is accurate and from a location close to where you are and you're not being sheltered from the wind by terrain etc... Anyway, it's fun as an experiment - more in the best bike split territory than real VE testing :) This works in the current iteration - if file has GPS data just check the "Wind Model" checkbox it will automatically calculate air density adjusted for elevation, airspeed and all that stuff.
Edit: one more thing, there's a setting in your MyCdA profile for drivetrain loss adjustment. Depending on what power meter you use that could also explain some differences in in the calculations.
Also crr might be a little higher than roller test data even if you're on a velodrome. There are some methods to determine crr using the VE protocol. For example, starting slow and increasing speed few loops so that in the end there's only a unique combination of crr and CdA that will level the VE graph. (
http://nyvelocity.com/...ess/the-shen-method/)
What's your CdA?