Tune test plotting seems off

[ksunden]

Here is a tune test taken on the PS table recently

This revealed two main deficiencies:

1. Our plotting seems to be doing something different than what we thought
   I've suspected as much in the past, but never saw it bad enough to be sure

2. When the colors are particularly far off (as with this where both ends of the curve were well off from the appropriate colors), we may actually want to accept the colors that the procedure fits, rather than remapping onto the original setpoints.

PR #68 deals with the latter, this issue is documenting the former

[ksunden]

Note that the grey line, should be the fits following along the maxima, but it is squished in this view so that its domain does not cover the values it should