Isn't there more of a window of chance that the heads could make contact with the surface/platter (or the media of the disk) if they weren't parked, thusly resulting in head sticktion or a head crash (I know the days of head sticktion are long over but is that partly because of head parking technology?)? I believe this was always more commonplace in notebook hard drives because notebooks are moved around with greater frequency (clearly), so heads are parked so as to avert disk damage/shock/vibration/collision. Of course, I don't know how often that could reach occurrence as I don't have that much experience with too many hard disks; from Western Digital, I've only had a WD5000BEVT (Scorpio Blue), a WD10EADS (a MyBook with a Caviar Green in it), and a WD6400AACS (Caviar Green). And I don't take a liking to the idea of having the hard disk heads park themselves on the ramp every 8 or 4 seconds either (the Caviar Green's timer is set to 8 seconds, whereas the Scorpio Blue's is 4, but the Scorpio Blue series has a minimum of 600,000 load cycles as the
spec sheet says). Maybe it would be ideal to set the timer to park the heads in a less abrupt manner, such as after 20-30 minutes of inactivity. Then again, I'm not sure how much wear head parking is on a mechanical hard disk as much as recording data to the media and writing/reading.
But I wouldn't want to make carte blanche assumptions. The only risk I could think of with head parking (that isn't in excess) is a shoddily cooled drive causing head deformation. But I think that's only an issue with drives that have heat issues. However, judging from the S.M.A.R.T. reports of the the first poster and second poster (worried about this issue), those hard disks have reached as high a temperature as 52C and 56C respectively (or 125-126F and 132-133F), and the critically hot range for hard drives is 50C (or 122F), at which point wear will be caused much sooner and bad/pending sectors are more likely to recur. Also, loading/unloading heads too often would also cause overabundant pauses in applications that load the hard drive often enough, such as watching videos or playing games, which is something to consider... furthermore, as also stated before me, it's important to ensure that the hard drive is not too cold (below 30C/86F under heavy load) as then it will have trouble lubricating and starting correctly, which may be much more harmful to the drive than occasionally warm or hot levels of thermal status, not unlike how a car engine will be more probable to die when started in cold weather. Avoiding thermal expansion is very vital as well (turning off/on electronics too much and too often, which is where the wear is greatest by way of great temperature fluctuations, for one thing). Anyway, I don't know how long those drive were at 52C/56C at. 39C and 38C are apparently much better temperatures. As for the Google study? It was done on drives ran on a 24/7 basis of usage (since they're server systems); for the sake of the electricity bill and eventual wear, I think many systems are ran closer to a 12/7 basis, which might nullify the study a bit (but not totally) and augment thermal cycling/expansion (and thereby wear and tear). Then again, that might not factor in standby or hibernation.
For the WD7500AACS and WD20EADS, however, on the spec sheet, as said before me, it says 300,000 is the
minimum rating for the load/unload cycle count, whereas the WD7500AACS in question more than tripled it. That isn't to put anyone at unease, though - minimum could even allude that a drive is capable of handling doubly that (over the duration of years, I would think, though, not one year). I guess a good balance to strike would be to set the idle timer to park the heads less often, but if 1,000,000 cycles is precarious enough then I don't think it would hurt to experiment and turn the timer completely off. There's a
bootable executable to do this called wdidle (
downloadable here), and the first link details the way to use it. But, I've heard that executable modifies firmware (voiding the warranty), and I'm not sure how well it works, so I suggest using it at your own discretion. However, Scorpio Blue drives' timer can be altered in Windows, through programs that can change the APM (Advanced Power Management) of the drive, such as HDDScan or quietHDD. There are others as well, of course, and 254 is the value to completely disable head parking, whereas 96 would be the default for Scorpio Blue drives, which is more of a risk than it is for desktop drives because values below 128 give way for the drive to sleep after inactivity and too many standby/sleep cycles for any drive will definitely cause premature death.
But it might be more in your comfort zone to tone down usage of your hard drive and see if the load cycle count becomes more moderate that way. As for the WD20EADS, its load cycle looks fine; to my knowledge, the WD10EADS/WD15EADS/WD20EADS series of hard drives are not as aggressive in their head parking. Also, sorry for the verbose and longness of this post. ^^;