FIX: section_distance calculation needs 3d support

[chrisbc]

We have a function solvis.geometry.section_distance() that doesn't account for dip angle or fault depth, so does not accurately calculate distance in all cases. It is not used but should be corrected and used.

Issues:

• test for shorter distances was failing on Mac/Windows but passing on Linux - Why
• we need some fixtures we can rely on for accuracy
• integration between shapely and Pyvista libs can be done getter (e.g. using get_geometry(include_x=)
• needs some visualisation of test outputs.

This issue was shown up by the #89