Bright light

Generality

It appears to apply to systems far-from equilbrium.

More specifically, it appears to apply from systems far-from equilbrium once they have developed self-organising systems in them.

Being far-from equlibrium alone is not enough - there are many systems which act as potential energy reservoirs, which are far from equilbrium - but are non-the-less quite stable - the maximum power principle does not usefully apply to them.

As such some of the most obvious systems it applies to are biology, systems involving turbulent fluid flow, drainage patterns and crystallisation.

Whether a dynamical system can be regarded as self- organising is often a matter of degree - and systems with "weak" self-organising systems in them may have the rate at which they consume resources governed by other factors.

There's also the question of whether all self- organising systems qualify. Looking at the main proposed mechanism by which self-organising systems consume their resources at a maximal rate, this depends on a selection process acting between variants - and it is conceivable that some self-organising systems don't fit into this framework very well. However, it seems to me that most self-organising systems actually do fit into this descriptive framework to some extent.