OK - lets try to explain better - all measurements approximate for the purpose of illustration (although they're pretty close)
As you can see from the below photo, the alloy disc centre has a recess - this is the side that faces the hub. The size of this recess is about 125mm
When I trial-fitted the disc, the hub flange was about 120mm in diameter. This means that with a PCD of 98mm and a wheel bolt diameter of 12mm, there was about 5mm of metal between the outer edge of the bolt hole and the outer edge of the flange - lets say this is the left-hand hub flange from this very-hastily drawn diagram.
This meant that the smaller flange fitted within the recess in the disc (and in fact, moved around a little, due to the slightly larger centre bore - see below)
When I came to fit the disc to my own car, I found that I had the flange on the right of the above diagram, which was simply a larger diameter - approx 128mm
This means that it has a little more metal between the outer of the bolt hole and the edge of the flange AND it means it was marginally too big to fit in the recess of the disc.
OK - that was problem#1
Problem#2 was that the flange boss (the bit that the disc and wheel locate onto) of the Coupe flange is 58mm. The centre hole in my disc is 59mm. When I fitted the disc to my smaller-sized flange, the disc clearly moved around (by 0.5mm in any direction), as the disc hole wasn't locating on the flange boss.
The original proposal was to take up this 0.5mm gap with a "filler-ring" - I was actually going to use a feeler gauge, cut to the correct length and wrapped around the hub flange boss. The disc doesn't put any weight on the boss (unlike the wheel), so the "filler ring" would only have been there for the purpose of keeping the disc concentric with the hub flange.
My new proposal is to buy two new hubs (on order) and have them turned down by about 1.5mm (ie 3mm off the diameter) so that they locate perfectly in the disc recess. These machined hub flanges will still have more meat on them than the smaller flanges I used for the trial-fit. As the machined flanges will still have a 58mm boss, there will still be a 0.5mm gap between the disc and the boss (which won't matter, as the disc will be centred using its OD, instead its ID). The wheel will still fit perfectly, as it locates on the original flange boss.
Hope this makes sense, but I take Szkom's point about possible weakening, so I'll check the hubs after a short period of use.
Does the above make more sense now?