The first place I saw these diagrams is in the first edition of Tom Cover and Joy Thomas's excellent textbook - copyright 1991. But the Venn diagrams are prese

The first place I saw these diagrams is in the first edition of Tom Cover and Joy Thomas's excellent textbook - copyright 1991. But the Venn diagrams are presented in their book as mnemonic tools, and not in terms of entropies of an underlying "I-measure" as in the article. (Of course, the book is an introduction, not in a theorem/proof format.) So you're not sure how far the notions can be extended. The world of information theory was small then, and the author of the paper would have been well-known to both Cover and Thomas.