Add diversity preserving samples reducer

Currently, one way of reducing the number of samples (in sampleset) is with the `SliceSamples` block (which acts only on one dimension, typically energy).

We would like to have a block that reduces the number of samples, but by eliminating "similar ones", and keeping relatively distinct ones in.

Start by reading the sampleset handling and the existing SliceSamples block. The issue does not name files, tests, or an entry point, and it leaves the similarity measure and diversity-preserving reduction behavior unspecified. Done would require an agreed algorithm and a block that reduces samples while retaining relatively distinct ones.

Written by the indexing model from the issue text.