ENH: Highlight countries with high share in addition to high number in international comparison plot

[corneliusroemer]

The international comparison plot for queries is super useful to quickly see where a mutation/lineage is most common.

Right now, covSpectrum autoselects around 6 countries to show. My suspicion is that these are the countries with the highest number of sequences. That biases towards countries that sequence a lot. While this is good for statistical robustness, it often misses countries that have a much higher share but don't sequence a lot.

It would be good if some countries with high share (not absolute number) were autoselected, too.