Very useful and insightful comments, flecc and trex. Thanks.
Taking into account your comments, let us KISS:
What about getting rid of the rear suspension / swing arm (which would be nice, but really not an integral part of my original idea) and building a hardtail (but still with a fold for the rear triangle)?
Then, the pivot of the rear triangle would not need to coincide with the BB, since the rear part would always be rigid (except when folding, of course). This would be similar to the rear folding of Brompton / Ori / Birdy / Airnimal Joey.
Would this simplification still face potential patent problems? I mean, are Brompton / Ori / ... paying for a patent by using that folding method?
Then, the remaining question is if it would be tolerable from a engineering point of view that when the fold of the rear triangle takes place, since the pivot of the rear triangle does not coincide with the BB, tension would be (slightly?) distorted. I remember watching a youtube video of the Ori suggesting to place the chain at the smallest sprocket before folding. Would this problem of the tension be a deal breaker, for a belt (I assume chains are more "flexible" regarding tensions)? Or it could be feasible?
To sum up, two questions:
1. If I get rid of the idea of BB = swing arm pivot, and I use a typical, non-concentric solution (such as Brompton / Ori ...), is there still a potential patent problem?
2. Is the use of a non-concentric solution a deal breaker for the tension of a belt (unlike a chain), when the rear wheel folds, around a pivot point which is not exactly the BB?