The paper cited does no such thing. Have you actually read it? It says twice in the abstract alone that it's a cloud-free model.Hence, the Kastings (1996) model, which does not take clouds into consideration at all, puts the furthest habitable zone radius for Kepler-186 at 0.2788 AU. Conversely, the updated habitable zone estimates (2013), which assumes H2O or CO2 clouds, places the furthest habitable zone radius for Kepler-186 at 0.4026 AU.
While the Kastings (1996) model relies solely upon the star's luminosity, the 2013 updated habitable zone model depends upon the cloud cover of the specific planet.