I am resurrecting this thread because I have finally obtained the technical details that I was after
Key facts:
1. Both the 7" acrylic dome and the 180mm glass dome have the same curvature radius so in terms of actual size there are no differences.
2. The 7" acrylic port has the centre of curvature 29.5 mm from the housing
3. The 180mm glass port with the 55 N120 N85 adapter and the 20mm extension ring has the centre of curvature 48 mm from the housing
4. The 8-18 entrance pupil at 8mm is around 38 mm from the housing
So in conclusion, once we remove the benefit of glass in terms of anti reflection and contrast in certain situations, both configurations are different from the ideal set up. Clearly the acrylic dome can't be adjusted as an 8.5 mm extension does not exist while the glass port could use a shorter extension of 10mm to position the dome in the ideal place.
There is however a post from @dreifish here that has used the 180mm with a 65mm extension saying that acrylic has an edge which does not seem to match any of the ideal considerations on dome positioning.
At the same time after speaking with nauticam they told me they use a macro slide to test MTF and the 75 mm length produces the best results. I do not have details on how they measure MTF and if they use corrected or uncorrected raw files however the 8-18 mm has a native barrel distortion of 6.2% according to optical limits and positioning the dome further away would introduce a pincushion distortion that would compensate the lens native barrel distortion and potentially result in a sharper image at expense of a reduced field of view.
In conclusion the improved sharpness of the glass would not have to do with the material being used but related to the different position of the dome. It has also to be said that the lens has less distortion at longer focal length and therefore there is a risk that pincushion distortion appears from a certain focal length on-wards that needs to be repaired in post and that the lens with the nauticam suggested extension does not perform consistently over the zoom range. Nauticam have confirmed they only test at wide end so this is plausible.
While it is not possible to extend the acrylic port of 8.5 mm it is possible to use a 17 mm mini extension together with a custom zoom gear (as the Nauticam one won't fit in the port) to bring the two ports at the same level of sharpness at 8 mm.
I am not convinced however that Nauticam approach is correct and I am inclined to suggest a shorter 10 mm extension on the port chart to let the lens perform as close as possible as it does on land and then do other adjustment to recover lost corner sharpness.