Evolutionary Tree

[Esquilax]

I understand the difference between fact and theory and that the word 'theory' has multiple meanings.  When I use them both in the same sentence however, my meaning is clear. So, when every third atheist tells me evolution is a fact, I should continue to remind them that only parts of it are fact. 

That really depends on what level of evidence you'd need to call something a fact. In my experience, people quibbling over that word, on this subject, are intending to apply an unreasonably rigorous standard for evolution alone, while significantly relaxing their standard for, say, their religious beliefs.

If you don't think that a thing that bears out in all of our recorded observations, and makes predictions that turn out to be dead on the money and would only work if that idea were true, is a fact, then I kinda don't know what to say to that.

Regarding comparing evolution to gravity, at least we can drop the apple to be sure (deduction) that it is a fact regardless of us not knowing how it works exactly.

You have an observation of a force, but not the cause of that force. How can you be sure invisible pixies didn't take the apple from you and push it down when you dropped it?

The problem is that you're happy to use the deductive, probabilistic method for gravity, which doesn't threaten your religious beliefs so much, but you're refusing to apply that same standard to evolution despite the much, much higher level of evidence in support of that theory. It's a double standard, is what it is.

However, in evolution, we have to infer from our observations that evolution is a fact (or not) -- and unless we figure out at least the major pieces, this inference is based on the assumption that the philosophical position of naturalism is true. I'm not saying this is wrong, just observing the distinction.  

Inference and deduction based on observation is the heart of science. You don't need to use "infer" like it's some dirty word; the fact that our observations- many of them so plainly obvious that they're a better candidate for deduction than your falling apple example- match that inference so consistently and over such a long period of time, even with the advent of new technologies and concepts that would have been unthinkable at the time evolution was first put forward, is a level of support in favor of that theory that gravity simply does not have. And yet you'll accept the one, and question the other. Are you just not aware of the constant, consistent reverification of evolution theory, or... what is the problem here?

And naturalism is not an assumption. All of our observations demonstrate that the natural world exists, and nobody has been able to do the same for the supernatural, even one iota. I know supernature is something you'd really, really like to be true, but you can't blame us for not accepting something out of hand based on no evidence, simply because of that. Do you understand how unreasonable you sound, when you assert that something that aligns with 100% of the observations is an assumption, merely for the crime of not entertaining something that itself has 0% of the evidence?

Regarding your closing remark 'Yes, because theories are comprised of constellations of facts. They are, in reality, much much stronger than a fact could ever be.' That would only be the case if there was a constellation of facts. Evolution seems to be missing a couple of core 'facts':

I'll go one by one, if it pleases.

common ancestry,

A deduction borne out by our consistently upheld observations that genetic similarity corresponds to ancestry, along with the fossil record showing said lineages and the predictions we can make, based on those two facts, producing accurate results.

how could a biological network evolve,

Gradually, over time, based on mutation and natural selection, something we've demonstrated even under laboratory conditions.

evolving traits with a low selection coefficient,

Randomized mutation and long periods of time, along with the obvious fact that a low selection rate is not zero.

convergent genetic evolution,

Certain genetic arrangements are so efficient for given sets of environments that they can repeat and persist, as there's nothing sufficient to drive them to change or select them out of the gene pool.

and more. So...how would you characterize a theory that is comprised of a constellation of theories?

I'd characterize evolution as a theory- in the scientific sense- that is complex and thus roundly misunderstood by those who argue against it.