[jschveibinz]

Some important things to consider:

1. Location estimates are noisy--you have to integrate over many samples to get a statistically significant result

2. Location estimates are 2-dimensional. The noise and statistics for lat are not the same as those for lon

3. Location estimates are dependent upon the spatial environment--radio signals are directional and they are affected by amplitude variations, antenna location, antenna visibility, changes in the immediate environment, etc.

4. Location estimates are affected by the quality and availability of reference signals and other factors like you mentioned: speed, orientation, changes in motion state (idle, walking, driving, etc), geofencing, etc.

You will need to include all of these factors in your testing to be able to determine the quality of your measurement tool.