My analysis:

SirHappyShoes is correct. However, CheeseOnWaffles gets an honorary mention for sounding like he really knows what he is talking about, even though he is completely wrong. He seems to know all of the terms, but has no idea how they really work. Perhaps he is deliberatly lying? Both contestents loose points for dumb names (I hope these are aliases used to protect the not-so-innocent and not actual nicks).

AxessTerminated:
Wouldn't every 2 of these cycles where the bits are all reversed bring the hard drive back to its original state?

1110010101 <--original
0001101010 <--after 1st cycle
1110010101 <--after 2nd cycle

and so on, so that after 6 it would be exactly the same data. Or did you mean that the first cycle would write all '1's, the second would write all '0's and so on?

Either way, you'd think it would be more effective to do six cycles of random bits, wouldn't it?

If you do the same operations to every bit on the drive, it would be easier to determine which ones had started off as '1's as they would be slightly more positive than the ones that had started as '0's. With random bits written in each cycle the forensic analysts would have to determine for each bit the last 6 random modifications before they could tell what the original data was.

Of course I know very little about the subject, and am just going on common sense. If you know why they do it the other way, could you enlighten us?