Strange behaviour of the Edge continuity tool in Rhino

There is a bug in Rhino which leads to an unwanted error of the reported accuracy when using the "Edge continuity" tool. When you compare the "Edge continuity" tool with the "Global edge continuity" tool on the same pair of edges, sometimes you may notice an obvious difference between the values they report. The "Global edge continuity" tool seems to be more accurate due to the higher precision up to 0,000001 digits, while the "Global edge continuity" is limited to just 0,0001 digits where the last digit is a rounded value.
This means that rotating a surface by 0,0005 degrees will be still reported as 0,000 degrees (perfect tangency), while rotating it by 0,00051 degrees will be reported as 0,001 degrees deviation, giving you a false assumption for the real values.
A more critical bug is the inaccurate graphs that change depending on the position of the edge pair in the 3d space relative to the CPlane origin. You will notice that the deviation graphs change every time, despite that the objects remain totally the same.