Better illustration of high-autofluorescence compensation models

[jakebeal]

In the presence of high autofluorescence, compensation models end up visualized poorly, as most of the data is effectively negative. Figure out a better way to illustrate them for this case.