improve calcParent2Child

[ralfHielscher]

the convergence of the function calcParent2Child should be investigated in more detail and improved

[nyyssont]

The ability to set the quantile would definitely help. Smaller parent grains would require a smaller quantile and larger PAGs a larger quantile.

[nyyssont]

Taking things further to the future... in some cases there might actually be two similar, but distinct representative p2c values in a dataset. Even in the example dataset there is some indication of this. Perhaps at some point a separate secondary p2c could be calculated for some misorientations that fit poorly with the main one?