OK, no one else has jumped in so I guess I'll take a crack at it. %-))
You haven't said how you measure your plates to confirm that they're actually linear. Nor have you said what lpi or screening that your using - both of which may help understand these numbers.
I think you have more problems than just the highlight bump.
Your M and C solid densities are too low - yet their final tone values are much too high.
Your K solid density is also way too low - yet it's final tone value is close to target.
Your Y seems fine.
I'll assume 175 lpi AM
If I were you I would start by exposing 4 identical plates (or expose one plate only) using the black screen only with a test form made up of tone values from 100% to 0% in 10% increments and 1% increments from 1%-10%. Confirm that the plate(s) are in fact linear and that all the dots are there - i.e. 1% and 2% and 99% and 98%.
The test form should look something like this:
You should also include a print contrast target to measure.
You print this same plate on each individual unit (or mount 4 identical plates on press). Print the colors individually - do NOT over print the inks. Run the inks up to the appropriate SIDs - which I believe are:
Yellow 1.15
Magenta 1.50
Cyan 1.40
Black 1.70
Get the densities even across the width of the press sheet and around the cylinder.
Then increase the SIDs by about 5% or whatever your dry-back allowance is. You are using a polarized instrument which adds another variable - but such is life.
Run at least 500 sheets or until your press operator feels the press is stable.
Then pull press sheets for evaluation. Each press sheet will only have one color on it.
The idea is that:
1) the plates are identical - that eliminates any plate imaging and/or screening issues/variables
2) there is no overprinting - that that eliminates any overprinting/trapping issues/variables
Examine the halftone dots and solids for each color using a microscope of at least 25x. You're looking for sharp-edged dots and solids (including the solids in the dots). You are checking for doubling, slur, ink over emulsification, voids in the solids, etc. The halftone dots and solids for all four sheets/colors should appear very similar.
If the dots are OK then measure your dot gains and plot your curves.
They should all have a similar shape and their dot gains should be close to correct:
Yellow 18
Magenta 20
Cyan 20
Black 22:
If you still have the highlight bump with one color then there's likely a mechanical or chemical issue with that unit or that ink. If your dot gains are still way out of range for all colors then you likely have an ink and or chemistry issue.
hope this helps, best, gordon p