Or someone's mathematical accounting is wrong. For example, the discovery of epigenetics changed things.The simple point is that the rate and nature of favourable (increasingly complex) mutations that has occurred in our biosphere hasn't been mathematically accounted for on a premise restricted to, and accurately reflecting, the foregoing two processes without a (3) third factor of some sort of bias towards complexity (or without appeal to a multiverse of infinite bubbles whereby this amazing coincidence can be explained as a feature in one -- our -- planet in one -- our -- universe).
first, natural environments are not in a "stable equilibrium", they're at best in a dynamic equilibrium (e.g. Lotka–Volterra predator–prey model), and climate/sea level history plus plate tectonics document external change that creates new ecological niches for living beings to evolve into. And environmental changes that open up new niches turn a bunch of "unfavorable" mutations into favorable ones. And complex dynamic systems exhibit chaotic behaviour simply because of their structure.Natural selection is a kind of purification process by which certain alleomorphs are purged from the population. This narrow process can never, even theoretically, account for the progressive complexification of lifeforms in the evolutionary process. In fact, without mutation, and once the effect of given selective pressures has played itself out, a closed population in a stable environment will quickly converge to a stable equilibrium state (Hardy-Weinberg) in which the proportion of all alleles is constant. Where no further genetic change occurs.
secondly, like a random walk, complexification isn't unidirectional. E.g. plants multiply their chromosome set to become polyploid, and use the redundance to simplify the genome (genome downsizing). We humans have a chromosome less than apes because sometime in our evolutionary history two chromosomes fused. An asteroid hits, or human civilisation happens, and huge numbers of species go extinct in the blink of an eye.
But there's a bias for complexity because becoming simpler is more difficult. It's unlikely for a random walk to not look complex, because the path contains its own history. Genetic mutations that add information are more likely to leave traits intact that are essential for survival, and adding information increases complexity.
I don't see any evidence that complex systems behave any different on Earth than they would elsewhere in the universe.