All of this assumes a given benchmark steel-taped distance with which to make the comparison, right? Maybe I should read the entire paper. Alan doesn't state how the "verified distance" was established.
What about frequency of reading? I know GPS units have improved significantly over the years. But to the extent that there are many fewer points recorded than a JC measurement, there will be some error regardless of the validity of the algorithm, no? Does anyone know what the reading frequency (recording of points frequency) current consumer models offer?