The first attempt certainly did have a hole in the green and blue, see post #36, top right O/P display.
The second attempt without reading errors, seems to have cured the problem as in post #39. The display has the advantage of being restricted to within AdobeRGB.
This shows an AdobeRGB spectral strip image top and a scanned print of the image bottom.
How much of the difference is due to the printer and profile and how much due to the scanning is unknown.
I suppose you now want me to do a print with the first profile to see if the holes really existed. I'll try to get round to it, but I am now looking at @pharmacist 's 750 patch target and puzzling out the file generation commands.
I have been experimenting for the first time, with creating @pharmacist 's 750 patch file. I have reached the stage of patch reading, hopefully I will be able to use my existing reader. The next stage could be 1040 patches
Later:-
I have read the 750 patches with the same reader, a couple of re-reads at the beginning, but otherwise without problems. I think 1040 is a possibility with the same reader using 26 x 40 patches.
I made the Pro9000II profile a 750 patch 6 ink by setting the paper to Canon Glossy Photo Paper, avoiding the use of red and green ink.
The Evaluation image file was converted from ProPhoto RGB to Adobe RGB and printed on Sihl glossy 280g. The image was printed using the new 750 profile and scanned.
I see you managed to produce 750 patch targets. Anyhow I also send you the set of the 3 tiX files to you as well. Getting 24x40 files on a single sheet of A4 is physically very difficult: the diameter of the reading eye is only 6 mm I think. The result is much better than the previous one: no strange dips in the bands with the different colors.