Justification is based on rather general grounds. Thin film is made of a particular material. If you model that bulk material using a given DFT functional (MGGA or just GGA, or whatever). It is a plausible procedure to adopt the same DFT functional when going to thin films or dots. Otherwise, how would you compare physical quantities of the bulk and thin film systems calculated with two different functionals? That would be like comparing apples and oranges.
Of course, if you are not interested in such a comparison, you may still tune c-parameter to get the electronic structure that you think is correct, but then you have to know what is correct. Using the MGGA functional with c-parameter from the bulk calculation actually allows you to make a parameter-free prediction of the electronic structure of thin films or whatever systems comprised of that material.
