@airwise: a good way to see what's going on is to use a steep long climb without wind to verify the power output. It is very important to know the exact system weight at the time,
to have no wind
Check - as much as possible. Slow speed. No headwind to speak of.
, and to use a barometric altimeter
or precisely measured course. I say steep and long because if you use a short climb (say 1km) any small errors in altitude data weigh heavily. Similarly on climbs that aren't very steep (say less than 7%) wind resistance takes a bigger role.
4km at over 9%. I'd say steep and long so check.
If your unit is drifting heavily then I would expect that the predicted and actual power start differing heavily. If this is not an issue they should be close. A word of caution, though: it is difficult to get valid data because of the limitations I mentioned above.
So we are simply supposed to "trust" our power meters are we? I ask because there is a lot of evidence to suggest that many drift enough to make the measurements fairly meaningless - certainly for the purpose of pacing a climb.
The climb gradient also needs to be very consistent in order to use VAM as a means of inferring power output.
Within reason - if averaged over a long enough period these things even out.