Say that the A-facts cs-supervene on the B-facts iff a computer simulation of the B-facts, based on the actual world's B-laws and initial conditions for B-facts, would necessarily also correctly simulate the A-facts.
Thesis: A-facts are problematic for physicalists iff A-facts do not cs-supervene on physical facts.
Corollary: There is a debate in the philosophy of biology on whether the apparent irreducibility of biological facts to physical facts is a problem for physicalism, and if not, why not. If the Thesis is true, then as long as biological facts cs-supervene on physical facts, then whether or not there is a reduction, biology does not present a problem for physicalism. I suspect that typical non-reductionist philosophers of biology would accept the cs-supervenience of the biological on the physical—for instance, they would agree that simulated natural selection would show up in the physical simulations—and hence their non-reductionism is not a problem for physicalism.
What about my argument against reductionism in biology, based on the instantaneousness of coming-into-being and the gradualness of chemical processes? I guess in the present setting the thing to say is that we wouldn't be able to tell from the computer simulation exactly when a new animal came into existence—just as we can't tell from empirical observation of the world—and so the computer simulation would not be simulating the coming into existence of animals.
A problem with the notion of cs-supervenience is the imprecision of the notion of "simulating".