>We still need to know what the thing is that the brain does

Yes, but not necessarily at the level where the interesting bits happen. It’s entirely possible to simulate poorly understood emergent behavior by simulating the underlying effects that give rise to it.