Scaling General Relativity

General Relativity (GR) remains the most successful description of gravitational behaviourwithin the dense, symmetry-stabilised region of the universe accessible to early 20th-centurymeasurement. However, GR was never designed to scale beyond that environment. Itsfoundational assumptions — constant light speed, smooth curvature, and universal symmetry— were local approximations, not universal truths. When these assumptions are extended intolow-density, gradient-dominated regions of the cosmos, they generate a series of observationaltensions: the dark matter problem, dark energy, inflation, the horizon and flatness problems,and the Hubble tension. These tensions motivate a more general framework rather thanadditional patch terms.The Canon framework meets GR precisely at the symmetry plateau where Einstein measured it, but extends in both directions: downward into recursion and coherence, and upwardinto density gradients and large-scale field structure. This paper argues that GR is completewithin its domain but cannot be universal. Scaling GR requires replacing symmetry withrecursion and replacing constants with density-dependent operators. The Canon framework isa validated theory grounded in demonstrated results across orbital mechanics, atmosphericdynamics, and classical physics. The variable-propagation extension to cosmological scales isa testable prediction of that theory, and specific falsifiable tests are identified.