In the "GPS Accuracy" thread, the issue was raised that the sampling rate itself can introduce speed errors, so I went and had a look at how much the top speeds differ if we look at the same sessions at 10 Hz, 5 Hz, and 1 Hz. Here is a graph that illustrates the measurement of a top speed at the three different rates:
The next graph summarizes the average differences, looking at the top 5 results from 10 different files in 3 GPSTC categories:
For comparison, the red bars show the average differences between sessions recorded with 2 different 10 Hz units from 5 sessions by 3 different speedsurfers. Here's the same graph for the maximum differences measured:
The measured differences between 5 and 10 Hz data are smaller than the differences between 2 units (which are already quite small, especially for 10 seconds and longer).
A more detailed description with plenty of numbers is at boardsurfr.blogspot.com/2022/01/measuring-speed-errors-sampling.html