Cross compare model skills

[Hendrik1987]

Capacity to cross-compare QI parameters of different models / model calibration cases. I.e., adding a dimension to the skill report.
There are many ways to visualize this, this could be one example:
Location vs. bias of Hm0 for various calibration runs of a SW model:

[Hendrik1987]

@jsmariegaard: After your demonstration yesterday, I now see that 'calibration' is already included as a use-case. So this idea is redundant, I leave it here for now for the example plot.

[jsmariegaard]

I think we should add some examples like this to a notebook. The data is all there by the skill() - but we should still show some good ways to visualize. Would it even make sense to do a Skill class similar to #41 for convenient plotting etc?

[jsmariegaard]

@Hendrik1987 this should now be "solved" with the plot_line() method in the AggregatedSkill class as of #47

[Hendrik1987]

@jsmariegaard: Great new features, yes, solved! 👍