What I've learned from things like this is that my intuition for when diffusion should work, and how fast, is just busted.
The biggest example in my mind is pyroclastic flows from a volcano. For years (mostly as a kid), I didn't understand what they were because I kept thinking roughly "if it was just gas it couldn't hold together into a coherent flow, right?" Well, as far as I can tell, yeah, they're basically just gas, but really hot and full of rocks. I still don't really get why they hold together instead of poofing out into clouds, except that it has something to do with how they're denser than the surrounding atmosphere. They're sometimes also referred to as "pyroclastic density currents", go figure. https://en.wikipedia.org/wiki/Pyroclastic_flow
Another big one that threw me for a loop as a kid was "warm/cold air masses" interacting in a way that kept their identity. "why does the warm air mass ride up on top of the cool air instead of just mixing?" I guess the answer here is that they're just too big for the mixing to happen faster than the bulk motion.
Anyway, it seems like fluids, especially fluids of very different properties, especially different density, take their time mixing, sometimes long enough to let them act like separate objects in contact for much longer than I, for one, would think. Thousands of years in the case of these brine pools (though in this case, they're being at least partially refreshed). I wish I understood it better.
PS: I was half expecting this EV Nautilus brine pool video. Grim comedy of a sort. https://youtu.be/9ZYJAmAmFPw