I don't know anything about the history of chemistry (because chemistry is balls), but if observations can't be consistent without a particular thing being true, those observations would count as empirical evidence for that thing.

So, let's talk about the Higgs boson. I'm sure we all remember how scientists had reason to believe it existed, but only 'discovered' it about a year ago. But let's talk about why they believed it existed in the first place. Physicists had a model for quantum physics, and it has very near-perfect predictive power and it explains many things. For this reason, it is considered to be a valid scientific model. One aspect of this model was that it involved the existence of a ''Higgs field'' which gives elementary particles their mass. However, the existence of a Higgs field implies the existence of a particle called a Higgs boson, with certain properties. Since this model is such a good one, this gave reason to believe that the Higgs boson exists, even though it had not yet been discovered. And then, of course, a particle with properties consistent with the Higgs boson was discovered. Going by what you said, the scenario is very similar in your example. Scientists had models which made accurate predictions about our observations, and these models also predicted the existence of atoms, therefore atoms were assumed to exist. So, to be clear: these assumptions are still grounded in empirical evidence, even if they may not have been conclusively proven; they are predicted by models which very accurately predict observations, and we therefore have good reason to make these assumptions. If you can assume God exists in a similar manner, I'd be very interested to hear your reasoning.

Oh no, I'm an atheist. I was just arguing the fact that sometimes in science, you can't always rely on empirical evidence to complete or validate a theory. Maybe someone can argue the existence of god given our examples, however I am not qualified to do so.