Improve coordinate concordance

[CBeelen]

The coordinate concordance used to be calculated relative to reference coordinates in 7.15, and will be updated to conseq coordinates in 7.16. However, both of these approaches have one drawback: we cannot calculate a concordance for insertions or deletions, respectively, because they have nothing to be compared to.
An alternative would be to extend both sequences to be the same length - that is, to fill up all deletions in both sequences with x characters - and to calculate the concordance for these two sequences. This would mean that the concordance decreases for insertions and deletions in the conseq equally - for a deletion, we would be able to tell because the window over which we average the concordance would extend into the deletion at its edge.
The advantage is that insertions and deletions would be treated "symmetrically". The difficulty will be to keep track of the true conseq coordinates and to assign the computed concordances to the correct positions. We will also have to figure out what to do with unmatched pieces of sequence. Currently, we do not compare them to anything and leave their concordance at 0.