I'm not sure I get what you mean... The graph also has a section of poly that doesn't overlap with open. A relation ship can be open and poly, open and not poly, poly and not open. It seems true to me. I fail to see why that graph is bad. Do you feel that "open relationship" should be a section that has no overlap with either poly or swinging? Because I disagree, personally.
Under the term "nonmonogamy," one can find poly folk, one can find, swingers, and one can find folks who are neither poly nor swingers who have open relationships. By trying to equate "nonmonogamous" with "open," one cannot use "open" to describe those relationships, because "open" then includes both poly and swinging, too, so the best one could do would be to say "it's open and not poly or swinging" which is one helluva cumbersome way to described things.
On the other hand, if one doesn't equate "nonmonogamous" with "open," then one can say that a relationship is "open" and it automatically means "nonmonogamous and not poly or swinging." "Open" thus refers to a sub-set of "nonmonogamous" and not the whole shooting match, which makes it a much more useful term; it also means we won't have to try to find a new term that means "nonmonogamous and not poly or swinging."
So, no, the meaning of open should not overlap with poly or swinging, as they are distinctly different things. People can engage in any particular mix of those things, certainly (and my wife and I are poly and open). The terms describe different things, though, in a useful fashion when that distinction is attached to them. IF the graph just shows all of the different ways people can engage in this thing and the other, then, sure, there's a need for overlapping areas. In terms of definition, however, it removes usefulness from the term "open."
