Vladimir,
I forgot to comment on the following your point:
"using classes seems an artificial top-down coarse approach to describe what in the end would be the most delicate operations of nature at the sub-atomic level."
As you may know, it is physics that approached nature in a "top-down coarse approach" and eventually met with the substantial difficulties in QM, where one has to deal with the "bottom" side of reality.
The ETS struct is a suggested bottom-up approach to the representation of "reality". However, what gradually became clear to me--and independently, to some extent, to some physicists--is that the basis of such approach inevitably leads outside the conventional spatial framework.
Physicists are not yet comfortable to think and to talk in terms of new forms of data representation, but this will come once you started on the path leading outside the spatial forms of data representation.