Auto Stop mechanism?

I doubt it uses the accelerometer since I experience false stops with auto-stop set to 1km, when walking briskly up and down switchbacking stairs like the ones you see on highway overpass bridges, and on the side of hills.

If GPS, what does the observed signal "drop-outs" suggest about the accuracy of stats, maps, etc that rely on GPS for data?

2221382 said:

what does the observed signal "drop-outs" suggest about the accuracy

It suggests that every technology has its limits. And that's so also for embeded GPS devices, especially in urban environment. If you are unhappy with the auto-stop feature, simply disable it, and pause the activity manually, when needed.

Of course the Auto-Stop option can be disabled but choosing to do so doesn't address the point of my question. Auto-Stop has revealed that the Instinct clearly drops GPS signal quite often over a course of travel - How does that technology limitation affect any related reporting for this device? Missing data points that create inaccurate analysis? Other side-effects?

2221382 said:

How does that technology limitation affect any related reporting for this device? Missing data points that create inaccurate analysis? Other side-effects?

Just have a look at the details of the concerned activity - zoom on the location, both on the map, and on the data graphs, and you should be abe to see what impact it had. From my own experience it is usually simply interpolated, but in some cases it may also lead to a wrong position, and hence to an incorrect path calculation.

BTW, the auto-stop is often caused simply by the GPS location fluctuation, and not because the watch would drop the GPS signal. GPS fluctuation is an effect where the detected position may fluctuate due to the limited accuracy of GPS in general (tens of meters for civilian devices), and/or due to limited reception.