Binary-coded decimal formats have more or less already lived and died (both fixed and floating point). They still have areas of applicability, but this idea is

Binary-coded decimal formats have more or less already lived and died (both fixed and floating point). They still have areas of applicability, but this idea is very much not a new one - x86 used to have native BCD support, but it was taken out in amd64 IIRC.