how much you can build without a complete understanding
We’ve never actually never had one. I’d have to check the timelines but Tesla was almost certainly working on a functional, but inaccurate atomic model (Bohr). Medicine is actually a great example of all this. We are so used to just kind of knowing “there’s a bad bug or bad gene that’s making me sick”. Like you may not know the details, but you’ve got some loose concept a bunch of cells in your body are pissed off. For the vast, vasssssssst history of medicine, it was all empirical, and the thing is, it kind of worked… sometimes.
My favorite example of “knowing without fully understanding” is Mendel and his peas. If you do a 4x4 punnet square (that gene cross thing), and look at the frequency of co-inheritance, you can track how far genes are from on another (because the further they are, the more likely there will be a swap during the shuffle). Thing is… because DNA is an integer thing (no such thing as ‘half a base pair’) it works DOWN TO THE SINGLE BASE PAIR. Mendel was accurately counting the number of freaking base pairs separating genes without knowing what a base pair, or indeed even really a molecule, was.
Tesla would have lived to see some absolutely nutty stuff in physics. Boltzman, Einstein with relativity, it must have seemed like pure madness at the time.
So yeah, we discover new and interesting stuff all the time. I personally think that some of the weird quantum stuff is going seem as rote in the future as germs do to us now. As in, the same way any lay-person shoved into a time machine would at least be able to give the basics to a medieval European, someone from the future would be like “well I don’t remember much about quantum tunneling, but…”.
And that’s all before getting into some of the bizarre things going on in math itself. Be careful if you look into that stuff though, it’s easy to fall into the “Terrance Howard” style rabbit hole. Suffice to say there is some really interesting and unexpected implications we’re discovering, but if you don’t have a solid grasp of theory, it is easy to be led astray but sources that want to gloss over details to talk about a conclusion that isn’t actually supported. It’s like if you tried to explain time dilation to an ancient Greek, and they excitedly hopped on their fastest chariot thinking they could “fast forward” to the future, because time moves “more slowly” for you when you’re going faster, right?
Somewhere, some patent lawyers are going to make millions debating about whether or not this constitutes “public disclosure”.