Based on GAIA EDR3, we revisit and update our sample of bonafide single stars in the Hyades open cluster. The small observational uncertainties in parallax and photometry of EDR3 result in a tightly defined stellar sequence, which is ideal for the testing and calibration of theoretical stellar evolutionary tracks and isochrones. We benchmark the solar-scaled MESA evolutionary models against the single star sequence. We find that the non-rotating MESA models for [Fe/H]=+0.25 provide a good fit for stars with masses above 0.85, and very low mass stars below 0.25M_{sun}_. For stars with masses between 0.25 and 0.85M_{sun}_the models systematically under predict the observed stellar luminosity. One potential limitation of the models for partially convective stars more massive than 0.35M_{sun}_is the prescription of (superadiabatic) convection with the mixing-length theory parameter {alpha}_ML_ tuned to match the Solar model. Below 0.35M_{sun}_, the increased scatter in the stellar sequence might be a manifestation of the convective kissing instability, which is driven by variations in the ^3^He nuclear energy production rate due to instabilities at the convective core to envelope boundary. For a Hyades-like stellar population, the application of solar-scaled models to subsolar mass stars could result in a significant underestimate of the age, or an overestimate of the metallicity. We suggest that future grids of solar-scaled evolutionary stellar models could be complemented by Hyades-scaled models in the mass range 0.25 to 0.85M_{sun}_.