In the traditional workflow, you just "print through" the thick neg, i.e. you do a longer exposure to the paper. You will also need to do a contrast adjustment, e.g. print to multigrade paper and fiddle with contrast and perhaps do split grade printing. This is because all the info you want is bunched up in a narrow range of optical densities.
In the scan workflow, the optical density of the thickest (=most overexposed) negative is well within what most scanners can handle. A good flatbed can manage close to DMax 4 - about as dense as Velvia can get. Even a wildly overexposed b&w neg reaches nowhere near that optical density. So what happens is you scan the neg and find that the tones are all bunched up, but then you 'auto-level' the tones to try and get the right contrast. It is very easy. Actually easier than traditional route, I daresay... but not necessarily better Anyway that is a topic for DPUG.
So... either way, the latitude of film real shines.
There are, of course, consequences of overexposure. But the point is that you can overexpose a neg film quite severely and still get a usable image. And even if you don't overexpose wholesale, another point is that little specular highlights will be held in detail, e.g. highlights on jewelry or a wedding dress, chrome on a car, reflections on water, etc...