The chemical enrichment of the Universe; the mass spectrum of planetary nebulae, white dwarfs and gravitational wave progenitors; the frequency distribution of Type I and II supernovae; the fate of exoplanets ... a multitude of phenomena which is highly regulated by the amounts of mass that stars expel through a powerful wind. For more than half a century, these winds of cool ageing stars have been interpreted within the common interpretive framework of 1-dimensional (1D) models. I here discuss how that framework now appears to be highly problematic. * Current 1D mass-loss rate formulae differ by orders of magnitude, rendering contemporary stellar evolution predictions highly uncertain. These stellar winds harbour 3D complexities which bridge 23 orders of magnitude in scale, ranging from the nanometer up to thousands of astronomical units. We need to embrace and understand these 3D spatial realities if we aim to quantify mass loss and assess its effect on stellar evolution. We therefore need to gauge * the 3D life of molecules and solid-state aggregates: the gas-phase clusters that form the first dust seeds are not yet identified. This limits our ability to predict mass-loss rates using a self-consistent approach. * the emergence of 3D clumps: they contribute in a non-negligible way to the mass loss, although they seem of limited importance for the wind-driving mechanism. * the 3D lasting impact of a (hidden) companion: unrecognised binary interaction has biased previous mass-loss rate estimates towards values that are too large. Only then will it be possible to drastically improve our predictive power of the evolutionary path in 4D (classical) spacetime of any star.