The situation described (cannot reproduce) is the standard in modern paper publishing, a significant fraction of even widely cited papers are just full of nonsense. The fact that it is not replicated in big publishers should therefore be sought in a different direction (I would say a political one).
