Okay, so what problem are you looking to solve? Right off the bat I see one print that's presentable albeit not optimal (#5); one is not so great because of the blown out highlights (#4). But I assume you initially had #4 and then worked towards #5 by means of improvement, right?
Right off the bat I'd say that part of the issue is that your inkjet negatives inherently block a little too much light. This results in the need for a fairly dramatic adjustment curve where you have to basically lop off part of the density range. This can result in unnecessarily crude tonality. The lack of contrast in the midtones in #5 is likely a result of this. It can however also be a problem in how the linearization process was performed (mixup of measurement data etc.) However, in principle, linearization can still work despite this issue, so you could simply accept it and move on. What kind of printer and printing software do you use? Did you make a blocker test that allows you to dial in which inks to use and how much ink density per channel (assuming these are configurable in your setup)?
Also, print #4 shows signs of poor contact between the negative and the printing paper; note the fuzzy part of the image around the man's chin. What kind of contact printing frame do you use?
In general, ensure that the kallitype printing process is
fully under control before attempting linearization. If your process still results in inconsistencies, you'll be chasing your tail forever trying to build a suitable curve. I'd start with a Stouffer step tablet and one known-good negative (inkjet or silver) and then practice printing that until the process is 100% reliable.
I notice that the print tone in your photos is very different from the density patches. Are these density and hue differences representative of the real prints?
(Screen Printing Exposure Light
Power:50W
Input
AC100-120V AC220-240V
CEE
B/N: K25A 1613SWO110400548)
That doesn't tell me much to be honest, except that it's on the lean side; 50W regardless of the actual type of light source is a little low. But it'll work if the exposure time is kept sufficiently long. Anyway, I see no signs in your prints to suggest there's anything wrong with the light source.