When CMB data are used to derive cosmological parameters, their very choice does matter: some parameter values can be biased if the parameter space does not cover the "true" model. This is a problem, because of the difficulty to parametrize Dark Energy (DE) physics. We test this risk through numerical experiments. We create artificial data for dynamical or coupled DE models and then use MCMC techniques to recover model parameters, by assuming a constant DE state parameter w and no DM--DE coupling. For the DE potential considered, no serious bias arises when coupling is absent. On the contrary, \omega_{o,c}, and thence H_o and \Omega_{o,m}, suffer a serious bias when the "true" cosmology includes even just a mild DM--DE coupling. Until the dark components keep an unknown nature, therefore, it can be important to allow for a degree of freedom accounting for DM--DE coupling, even more than increasing the number of parameters accounting for the w(a) behavior.