Originally Posted by mac
One of my goals is to get a nice road bike geared high enough where I'm not spinning at 120 rpms and bouncing my butt in the saddle and fly down the mountain at 50 mph. I use both a GPS and cyclocomputer, but trust the GPS more. There's a mountain pass highway as you exit Van Nuys north into Santa Clarita where I got up to 44 on my commuter. Now that was a rush!
For measuring an instantaneous top speed I would rely on a calibrated computer long before a GPS measuring system. There is too much individual sample error in GPS units to get an accurate single point or peak measurement (i.e the difference between two samples), for averages they work much better as you are using many more samples and measuring greater distances, so a few feet of error becomes very minor.
Al