And then I thought I might see what the difference is between Strava's estimate (used the app on my phone) against the Quarq's actual power data. This is from the headwind section above.

First the Quarq data as shown on Strava:



And then the Strava estimate of the same segment:



Strava estimate: 167w
Actual data: 306w

+/- 139w is not what I would consider in the ballpark. But it really points out the affect that wind and aerodynamics have on cycling. And the challenges you'd have designing an app that could reliably estimate power without actually measuring it.