I don't think that abstraction is particularly useful because beyond extremely simple systems, approaching things from physics becomes 1.) not possible because

I don't think that abstraction is particularly useful because beyond extremely simple systems, approaching things from physics becomes 1.) not possible because we don't actually know enough and/or 2.) intractable because there are too many variables/actors at play. You can get insight from physics, but physics doesn't actually help you much when it comes to actually making a molecule, or figuring out how to avoid reacting with one part of a molecule while selectively modifying another. In other words chemistry is to physics as computer science is to math. They're related, you could say one is reducible to the other, but that's not a useful construct in practice.