I see, we worked out what was meant.

Strange enough, sometimes opening the LS needle until it is no longer active, ALSO works.

On second thought, I see Chris' curve ends at around +40 on a -100 to +100 scale, and that is a bit low...

Now there are two different trains of thought here. If HS mixture is largely determined by the solenoid (relatively wide open HS needle) and idle mixture is on LS needle AND solenoid, that causes the dip in the curve.
The location of the lowest point of that dip is determined by how many points there are in the curve, and whether those points are also X-axis adjustable or not. That lowest point is NOT necessarily where it NEEDS to be according to the engine.
That will always cause a bit of "friction" between engine requirement and curve possibility. Because MAYBE the engine wants a slightly leaner mix where the curve, because of its fixed X-axis, allready rises or vice versa.
The way around this is to open the LS needle to where it is no longer active, and do most of the WOT mixture on the needle, That will end up in a curve that is always ascending over the entire range, and that eliminates the issue of the lowest point not coinciding with where the engine wants that lowest point.

So what I would suggest, IF "closing the LS needle to force the curve up" does not fix the issue, is to open the LS needle a few turns (it should not fall out but also be not active anymore) and close the HS needle to where the solenoid just comes in (appr 70~80 on a -100 to +100 scale) in order to create as much "range" as possible, and run the entire mixture on the solenoid.

I have a few engines where this seems to work pretty well. Idle mixture most definitely is less refined, but it cures the midrange issue most effective. And since I only idle my engines on the ground, I do not care too much for that.
I think that stuff like "the engine has to idle low enough to not roll on pavement" highly overrated. It has to idle low enough for a reliable approach and landing. That is all I need.