Race all the way down, race all the way up: A substrate-universal race-architecture across quantum, classical, biological, and computational regimes

Scope note: Friction Theory (FT, Pødenphant Lund 2026b) formalizes bounded probabilistic computation systems satisfying the race-axioms. FT's established scope covers biological, cognitive, and computational substrates (Papers 1–6). This paper investigates whether FT's mathematical structure scales to physics-scope substrates (quantum measurement, classical mechanics, thermodynamics). We argue that "decision" in resource-bounded probabilistic systems is not a mental or agentive primitive but a substrate-universal structural phenomenon: the resolution of competing processes racing toward commit under a finite-time budget. Any system satisfying R1 (parallel candidates) + R2 (bounded resources) + R3 (irreversible commit) exhibits an inverted U on its evaluation-to-commit rate — too low yields no information processing, too high yields noise-dominated commit, and only the intermediate rate maximizes information throughput. Seven apparently independent phenomena — qubit decoherence-window, Ohm-Drude electron transport, chemistry/biochemistry molecular kinetics, stochastic resonance, Margolus-Levitin saturation, encoding-friction in learning, and Yerkes-Dodson — are manifestations of this single structural necessity across substrate scales spanning forty orders of magnitude in characteristic timescale. We refine R1-R3 to a five-axiom formulation A1-A5 appropriate for Schwinger-Keldysh derivation, and demonstrate that A1-A5 are satisfied by the closed-time-path generating functional of any bipartite quantum system with einselected pointer basis and Markovian environment. As corollaries we recover the Feynman path integral, Onsager-Machlup stochastic dynamics, the Friston free energy principle, and a CR-signal in large language models as parameter-regimes of one underlying derivation. Time emerges from commit-sequences. We engage interference, linearity-nonlinearity, and reversibility honestly: the framework relocates rather than solves the measurement problem.