> Don't some of them show your speed as well as location?
I've found that the accuracy of my older Garmin GPS 45 (even with SA turned off) is dependant on the number of satellites visible and the amount of time I'm willing to give it to grind through the numbers. Moving through timber and through canyons limits the visible constellation (of satellites). On back country mountaineering trips, the GPS is frequently useless for hours at a time because I can't find enough open sky to get a fix on more than 2 satellites. The flatter, and more open, the country, the better a GPS works.
Back when SA (Selective Availability, a deliberate encoding error introduced to limit the accuracy of non-military GPS units) was turned off, another guy in my office and I walked over to the King Dome (Seattle WA) parking lot at lunch to see how things would go. We had an accurate location from a relatively recent USGS map. After 10 minutes, with 6 satellites visible, the two units converged on a location. They agreed on all but the last significant digit (we were using UTM, not Long and Lat). The last digit they took turns bouncing on (5, then 6, then back to 5. Both units.)
To get back to velocity, or speed... it's calculated (as I understand it), on how long it took you to travel from point A to point B... it doesn't necessarily take into account how far you travelled to get there. It's an "as the crow flies" calculation.
So... I don't think GPS would be worth the extra expense over bicycle speedometers.
" />