They did it because newbies keep installing hard drives and wondering where some of their space is, "I bought this 320gb and it's only giving me 298gb".
Bllsht! It used to make sense 20 years ago to build 1024 byte RAM cells instead of 1000 byte RAM cells. And it also used to make sense to call such a RAM cell a one kilobyte RAM cell. What's the advantage of this arbitrary choice for hard disks? Their sizes never aligned to power of 2 sizes anyway. Their capacity was limited by the number and size of platters and their density. Their were not build in blocks of power of 2 sized elements.
If you are into computer science you will be certainly able to tell? Which calculation, algorithm, or alignment issue would be at all affected if kilo, mega, giga, and terabyte sizes for hard disks would be counted in base 10 increments?
Oh, I see there's actually already the mentioned 9 page thread and this 3 page thread, where I have just posted a more detailed answer. Hardly "unnoticed".