The paper's leitmotiv is condensed in one word: robustness. This is a real hindrance for the successful implementation of any multigrid scheme for solving the Navier-Stokes set of equations. In this paper, many hints are given to improve this issue. Instead of looking for the best possible speed-up rate for a particular set of problems, at a given regime and in a given condition, the authors propose some ideas pursuing reasonable speed-up rates in any situation. In (Vázquez et al., 2001) (Vázquez, 1999;Vázquez and Codina, 1998).