A production runtime should have timeouts, memory ceilings, cancellation, and emergency shutdown. Those controls protect the host. They do not, by themselves, demonstrate the emergent structural halt defined in Teleodynamic Learning.
What the stronger claim requires
Noop must be a scored candidate. It must win repeatedly because structural alternatives no longer justify their predictive, complexity, and resource terms. The structural transition rate should collapse while parameter updates remain possible.
Control the confounders
Record every external cap and run experiments where those caps remain non-binding. Compare the result with a hard-stop baseline and with resource coupling disabled. Otherwise, “emergent” is only a label applied after an ordinary stop.