Improve density fit plot

[colinpalmer]

Moorhen's density fit plot (in the Validation plot panel) is giving me some unusual results. Here's an example of a model fitted with Molrep where the last few residues don't fit the map density:

And the same map and model in Coot 0.9:

These residues are basically in the noise so I assume the map density is below zero in places, and Coot 0.9 (correctly, IMO) shows that as a poor density fit while Moorhen shows it with a negative bar instead, which is much harder to interpret.

A few comments and suggestions:

• Could Moorhen show poorly-fitting residues with bars that all go in a consistent direction, regardless of whether the map density is above or below zero? (Or use colouring to make it clear that large bars in either direction are both bad!)
• It's hard to understand the scales used in the Moorhen validation plot. In Coot 0.9 the red-green colouring makes it very clear that a large bar indicates a poor fit. In Moorhen, I'm not so sure. Do large bars always indicate a problem or are they actually a good thing for some of the metrics? Could you add a guide to interpretation? Either something visual on the plot, or maybe a help panel that the user can open up to explain what the plots mean?
• An option to rescale the plots would be helpful, e.g. if there's a large bar that makes all the others so close to zero that they can't be seen
• It'd be nice to have a way to view the raw map density values for a particular residue - i.e. take the same map density information that's used internally to calculate the density fit, and make that visible somewhere, maybe as labels on the atoms or entries in a "residue info" table