WBS: LAS file resolution

[magnesj]

Based on mail from @lasserenli dated 25.05.2020

Resolution not taken from LAS file when LAS is used. Thus, high resolution features in the LAS file not being used. Should always re-sample to smallest increment or be possible to select reference for sampling in reservoir and non-reservoir separately.

[magnesj]

Related to #5602

[lindkvis]

We need to resample no matter what, because if we want to sample from a LAS-point, this has to be combined with some grid-data in the WBS-equations. Now, if the LAS-file has a worse resolution than the grid, we would risk making things worse if we always used the LAS-resolution instead. So I think we have to combine them. Also, we can't really just resample to the same increment as the LAS, as we still might miss the LAS-point. Also, multiple LAS-files may be used and they may have different points.

The most sensible thing to do I think, is to create a combined path of grid intersections and all LAS-points and use this for the sampling. Since grid-indexes are used to get locations currently, we would require a new data structure that can hold also hold LAS-points. Perhaps a vector containing well path point interface-objects that can be either LAS-points or a grid intersections, and for each one of them we make a judgement about how to get the data.

At the same time we should consider getting rid of the OpenMP code in the extractor for almost anything. It is likely too small and too few sub-tasks to provide any benefit.

[magnesj]

Fixed by #9955