Improve user-friendliness for incremental datasets

[sbrugman]

Incremental datasets can rely on a batch ID and stitching of histograms. Compared to working with time slices, the incremental dataset is much less user friendly:

• The labeling of histograms is awkward (the bin center is 1.5, 2.5 etc.)
• The documentation for entry-level usage (i.e. a single column with a "batch_id" column) is lacking.
• The pm_stability_report docstrings/arguments are tailored towards time-slices.

This user experience for this way of using popmon should be improved.