From what information I was able to find, I think what happens to OP's Genelec is probably "within the design intents". To have 2 concentric parts fit together and have a visually even small gap between them, you'll need very tight assembly tolerance and an assembly that is designed with that in mind. Here we are probably looking at a nominal gap width of a fraction of 1 mm, and to have a gap that is visually even all around, you'll need both parts to be almost perfectly circular and concentric to within a very small fraction of 1 mm.
Here is a picture of an example of the kind of issues design engineers typically have to deal with.
View attachment 535457
Now, to the actual design of the part and assembly ... Here is the cross section of a Genelec coax driver. Looks like the mid-range driver cone is composed of a foam membrane bonded to a (smaller) thin rigid cone at the back. The foam membrane is bonded to the outer driver frame at its outer edge and the foam membrane. The foam membrane is allowed to flex at its outer edge (between where the thin rigid cone ends and where the foam is bonded to the outer frame of the driver) and functions as the "surround".
Notice that the location of the outer edge of the mid-range cone (the foam part bonded to the frame) is located by the bonding process and is probably not very tightly controlled.
(Picture source:
https://audioxpress.com/article/r-d-stories-genelec-8351-acoustically-coaxial-sam-system, annotations by me)
View attachment 535459
The way the coax driver is located to the speaker enclosure is actually designed for good concentric alignment, with o-rings to "decouple" the driver from the enclosure. However, because the outside edge of the surround is only located by the bonding to the driver frame, the concentricity between it and the circular opening of the frame isn't tightly controlled, and therefore, IMHO, the Genelec design does not aim to ensure, and is not adequate for ensuring, that the gap between the driver and the enclosure opening is visually even all around. Since the wavelength of 20 kHz sound wave is 17 mm at 20 °C, a concentricity mismatch of a small fraction of 1 mm will not matter.
Acoustically, I believe how well the driver edge is flush against the enclosure is far more critical (and therefore the mentioning by Genelec of the "diffraction-free" joint between the MF cone and the "DCW™" – Genelec's name for its "directivity control waveguide").
(Picture source:
https://www.fast-and-wide.com/images/stories/White_papers/genelec 8260a - public white paper - v1.pdf)
View attachment 535460