cosmology works?

a freely-falling observer experiences the metric \(g_{\mu\nu}\) of generated by any energy-momentum distribution \(T_{\mu\nu}\) as being of Minkowskian shape \(\eta_{\mu\nu}\). the observer would conclude that there is no curvature and consequently, no gravitational fields and that the energy-momentum is vanishing. but isn’t that what we’re doing in cosmology? as a freely-falling observer we aim to determine the energy-momentum of all cosmological fluids \(T_{\mu\nu}\) expressed in terms of their density parameters \(\Omega\) and their equations of state \(w\). but how can one measure the energy-momentum-tensor of a gravitating matter distribution with a gravitational experiment?