@martinthomson wrote:
Bonus question: What is a "position accuracy"? In metrology, the area this described might be the 2d analogue of a confidence interval at a given confidence level (95%, 67%, something else). But the word "accuracy" implies something else.
We should improve the definition of that.