Marco Santambrogio, "Meinongian Theories of Generality," Nous, December 1990, p. 662:
. . . I take existence to mean just this: an entity, i, exists iff there is a determinate answer to every question concerning it or in other words, for every F(x) either F[x/i] or ~F[x/i] holds. The Tertium Non Datur is the hallmark of existence of reality. This is entirely in the Meinong-Twardowski tradition.
In other words, existence is completeness: Necessarily, for any x, x exists if and only if x is complete, i.e., satisfies the property version of the Law of Excluded Middle (Tertium Non Datur). Now I have long maintained that whatever exists is complete, but I have never been tempted by the thesis that whatever is complete exists.
Why can't there be complete nonexistent objects? Imagine the God of Leibniz, before the creation, contemplating an infinity of possible worlds, each of them determinate down to the last detail. None of them exists or is actual. But each of them is complete. One of them God calls 'Charley.' God says, Fiat Charley! And Charley exists. It is exactly the same world which 'before' was merely possible, only 'now' it is actual.
So why should completeness entail existence?