What do people think of this paper? If I understand correctly, they are saying that the need for Λ disappears if we do N-body simulations that properly take account of GR. I think they are saying that past simulations have used Newtonian approximations that don't properly account for density fluctuations. Can this be? If true, it seems like a major advance. Comments?