As far as I am aware the only way to remove outliers on a cloud to curve deviation is via a VB script or external script. The idea would be:

- Best fit cloud to curve

- Report 3D deviation for each point, and remove based on standard deviation

- Optionally iterate a second time to best fit without outliers included

Again, we currently solve the issue with a VB script that loops through the report items and doing exactly what I described, but a built in feature would be both convenient and more robust.