Thanks Peter,
Between this sample and a post by @DanielPiker
I was able to match the absolute curvature deviation from the EdgeDiscontinuity command as described here
By extracting the principal curvature from the osculating circles and converting it into a Radius value that aligns with the commands results.
Then using
|(1/Radius1)-(1/Radius2)|
to get the absolute deviation.
I am still hung up on getting the relative value using the formula
(Radius1-Radius2)/(Radius1+Radius2)
as the results do not match what EdgeContinuity command is returning.
I am sure I am doing something wrong in the process and any guidance would be appreciated.