Sigh. Oxygen has properties that have nothing to do with fire. You need it to properly model cellular respiration, water electrolysis, air currents, buoyancy, the properties of compounds of which the element is a part etc. Give me a coherent periodic table of elements that includes phlogiston instead of oxygen and we can talk.

Every physical theory we've come up with, when examined close enough, is completely and utterly wrong. If we're going to have any useful definition of accuracy, you can't just throw it out of the window because of that.

Some theories are less wrong. So yes, you absolutely can throw a physical theory out the window if it is wrong. You might save the equations so you can make quick, approximate calculations (i.e. Newtonian mechanics) but that doesn't mean you include all the entities in the theory in your ontology.

It worked perfectly for almost everything they did at the time.

This is essentially a truism for all outdated scientific theories.

For that matter, it works perfectly for almost everything we're doing now.

Sure unless you want to make sense of combustion and anything that requires knowledge of modern chemistry or atomic theory at the same time!