The increasing precision of spacecraft radiometric tracking data experienced in the last number of years, coupled with the huge amount of data collected and the long baselines of the available datasets, has made the direct observation of Solar System dynamics possible, and in particular relativistic effects, through the measurement of some key parameters as the post-Newtonian parameters, the Nordtvedt parameter η and the graviton mass. In this work we investigate the potentialities of the datasets provided by the most promising past, present and future interplanetary missions to draw a realistic picture of the knowledge that can be reached in the next 10-15 years. To this aim, we update the semi-analytical model originally developed for the BepiColombo mission, to take into account planet-planet relativistic interactions and eccentricity-induced effects and validate it against well-established numerical models to assess the precision of the retrieval of the parameters of interest. Before the analysis of the results we give a review of some of the hypotheses and constrained analysis schemes that have been proposed until now to overcome geometrical weaknessess and model degeneracies, proving that these strategies introduce model inconsistencies. Finally we apply our semi-analytical model to perform a covariance analysis on three samples of interplanetary missions: 1) those for which data are available now (e.g. Cassini, MESSENGER, MRO, Juno), 2) in the next years (BepiColombo) and 3) still to be launched as JUICE and VERITAS (this latter is waiting for the approval).