Electromagnetism, to take your example, is very much a theory that we can use to make predictions about a new system once we can write down its description. Yes

Electromagnetism, to take your example, is very much a theory that we can use to make predictions about a new system once we can write down its description. Yes, we sometimes make new discoveries and we have to develop new theories for those (like when we discovered that electricity and magnetism are governed by the same force), but that doesn't mean that for every new circuit layout we have start afresh by collecting data. We can apply the theory to the description and get a reasonable prediction. And that's what feels different about organic compounds. We already have the formula, but we still can't deduce much from it.