Hello Joe --
I'm really glad you enjoyed the essay; thank you.
Physicists like renormalization when theories are (ahem) renormalizable--meaning that there's decoupling, or simplification, in the IR (or, in the case of asymptotic freedom, UV). Here I don't think we need anything like that. It's OK to have a rough and jumbly effective theory that works only partially and can occasionally collapse or blow up.
I guess I'm asking is our perception the effective theory or is classical mechanics the effective theory? Aren't they both? And does that mean that our coarse-grained experience of the world somehow captures a glimpse of aspects of nature that are nevertheless causally separated from the low-energy descriptions that must hold in order for biology to emerge in the first place?
Both; we have plenty of effective theories that work at different scales and domains. And, indeed, I'd agree: we are sensitive to causal properties that do not appear at the most fundamental level of description. At one and the same time, X has causal power, and X does not (indeed, can not) appear at the microscopic level. Like (for example) entropy, or enthalpy, something that tells you which way the reaction is going to go without being related to any basic property.
By the way, I think people get confused about this. People hear that the fundamental laws of physics are deterministic (for example), and they worry about free will. So then they say, OK, I have to get rid of determinism at the fundamental level, and we get these crazy arguments about quantum mechanics and free will which I think are both wrong ("quantum" randomness is not special, and the wavefunction evolves deterministically), but more importantly really just missing the point. We don't need to bake X into the microphysics to get X at the macrolevel. (I think you agree with all this.)
Yours,
Simon