Assume that the start state is not the single final state. Suppose M is a lambda-NFA with exactly one final state f, but not necessarily satisfying any of our other assumptions. Make a new lambda-NFA M’ by adding two new transitions (f, lambda, i) and (i, lambda, f), where i is the start state. Is L(M’) always equal to L(M)^*? Either prove that it is or give an example where it is not. Does adding our other assumptions change anything?