A conjecture concerning a limit of non-Cayley graphs (Q5949012)

Graphs \(G\) and \(H\) are quasi-isometric if there exist a function \(\theta:V(G)\to V(H)\) and constants \(C, D\geq 1\) such that for all \(x,y\in V(G)\) and all \(z\in V(H)\) the usual distance metrics in \(G\) and \(H\) satisfy: (1) \(d(\theta x,\theta y)\leq Cd(x,y)\); (2) \(d(x,y)\geq D\) implies \(d(\theta x,\theta y)\leq d(x,y)/C\); and (3) \(d(\theta[V(G)],z)\leq D\). Intuitively, \(G\) and \(H\) ``look the same \(\ldots\) from far away.'' \textit{W. Woess} [Discrete Math. 95, No.~1-3, 373-384 (1991; Zbl 0757.05060)] posed the question as to whether every vertex-transitive graph is quasi-isomorphic to some Cayley graph. The present authors construct a sequence of graphs ``that seem to look less and less like Cayley graphs.'' The sequence converges, in a certain sense, to a 5-valent, 1-ended graph which is conjectured to be a counterexample to Woess' query. Reviewer's remark: The volume and paper number of the cited reference by W. Woess are incorrect in the article. The correct numbers appear in this review.