With some 14, 15 and 29 years of HMI, MDI and GONG observations, respectively, we have reached a new milestone in acquiring almost three decades of nearly continuous helioseismic observations. Moreover, an alternative fitting methodology to the respective projects global helioseismic pipeline has produced tables of individual mode frequencies based on long and very long time-series, namely from 3.2 to 12.6 year long time-series for all 3 instruments to even a 25.2 year long one for GONG observations. Having fitted such long time-series results in having very precise individual mode frequencies and individual rotational splittings. Hence we ought to revisit our traditional rotation inversion techniques and methodology to derive the best estimate of the solar internal rotation. To achieve this, we have reassessed some of our preconceived conceptions, like the density of the model grid used for inversions, but also developed a new iterative technique to complement the classical regularized least square one. We present new results from inverting rotational frequency splittings derived from these long and very long time-series, with an emphasis on optimizing the resolution near the tachocline. We show how we validated our inferences using simulations and, of course, recognize that our inferences correspond to the mean rotation rate over the extent of the fitted time-series.