[vtkusers] inconsistency in polydata connectivity filter

[Dominik Szczerba]

I run the polydata connectivity filter with 
SetExtractionModeToLargestRegion. It removes the small disconnected 
regions as expected, but does not remove the points contained by the 
deleted cells. It does remove these hanging points, however, on the 
subsequent call to the same method (using a deepcopy of the previous 
result as an updated input). Well, nothing crashes (yet), but it does 
not seem clean to me. Am I wrong?

-- Dominik