Surface using of linear interpolation on the delaunay triangulation

[MichaelClerx]

@mirams says:

Voronoi is sort of pretty, I guess the linear interpolation on the delaunay triangulation would be closer to the underlying function (piecewise linear rather than piecewise constant). The delaunay is 'dual' of voronoi so should be computed in the same time or as a pre-requisite for voronoi (we used to do this stuff in cell based Chaste)

This would be fun to have in addition to the voronoi surface plot. Might be cool for someone to add if this is easy