Give me a young star, and I can use the reductionist laws of physics to predict that star’s future: It will live a million years rather than a billion years; it will die as a black hole rather than as a white dwarf. But the components of a living organism yield something new and unexpected, a phenomenon called “emergence.” Give me a simple cell from the early days of Earth’s history, and I could never predict that some 4 billion years later it would evolve into a giant rabbit that can punch you in the face. Kangaroos—like humans—are an unpredictable, emergent consequence of life’s evolution.

But couldn't "reductionism" simply chalk this inability up to the practical lack of physical information and our purely technical inability to play out (simulate) something as massive as, at the very least, all that takes place on an entire planet?

In fact, even the example with the predictability of a star's life isn't all that certain in practice – do we really know exactly how the star's life will play out, or just generally? Will the Sun become a red giant in roughly ten million years, or in exactly 10.285.914 years? It's still a complicated chaotic system and we certainly can't account for all the details and microflunctuations. The same inability applies to physics describing evolution, with the main difference being how far-reaching the difficult-to-predict micro-flunctuations can be (a change in a gene can change everything about life on Earth millenia down the line; a solar flare, while involving an incomprehensibly larger amount of energy, changes next to nothing about how the star's life will play out, as far as we're concerned).

I take these differences more as a spontaneous consequence of how you frame your topic of study, depending on your practical possibilities (different methodologies arising based on how suitable/doable they are for different objects), rather than as a strict border between hard determinist physics and non-physical magic.