Separate chi² and arclength thresholds
Created by: Zaharid
There is no logical reason why these two should be the same (or it is rather weak). Make that explicit in the code, and allow to run quick and dirty tests.
This is the first step towards #527 (closed).