A recent scientific study at the University of California has discovered the process by which a hard drive magnetically fails. The solution is, somewhat obviously, to use better magnetic materials in drive platters. While this discovery could lead to better quality hard drives in the future, it won't solve other drive failure types. Maybe we won't even have to worry about this for very much longer if light replaces magnetism as a data storage and retrieval mechanism.
Ultimately, the goal in all electronics development is to replace the slow -- magnetism, physical switching, mechanical connections, and electrical signals -- with the fast, which generally involves switching to some sort of light-based solution. Fiber optics and light signals, for instance, will need to replace metal wire and electrical signals if we hope to improve signal bandwidth, speed of operation, reliability of connection, and greater power efficiency. So in essence, this discovery about hard drive failures comes too late in the magnetic hard drive life cycle to make a big difference. Magnetic storage is dying a slow but certainly deserving death.
Platter polarization is not the only way for a drive to fail, though. In fact I'd say in world of hard drives, it's probably the least likely to be at the heart of a catastrophic hard drive failure. Most common are the usual suspects -- electrical and mechanical failures. One of the realities of electro-mechanical devices is that even if the electrical components magically last forever, a lack of adequate lubrication and normal wear and tear will eventually cause an unrecoverable mechanical failure. Anyone who has owned more than one motor vehicle knows this reality all too well.
Beyond their initial "expect to fail" period of 30-90 days of regular use, electrical components rarely fail under normal conditions. It takes some kind of abuse, contamination, or corrosion from environmental factors to cause late-life electrical components to die. With hard drives, the extreme heat generated inside of your computer (much of it by the drive's own mechanics) and the thermal stress of turning it on and off on a daily basis is enough to cause both electrical and mechanical failures. Fortunately, an electrical failure is usually very easy to fix on a hard drive -- you just buy an identical drive and swap out the control boards.
Mechanical failures are by far the most common in my experience as a technician. Lack of lubrication can cause "hard drive sticktion," which causes initial failures at boot time, but can go away after the drive motor has had the chance to spin up a little. The only "voodoo" way of fixing this was to shake the drive vigorously. Sometimes the heads just don't seek anymore, and like a record with a dull stylus, the drive is unable to locate and read the requested sectors. Under these conditions, repair is not usually possible. There are so many problems with electro-mechanical magnetic storage; I can hardly wait for my first light-storage drive.
Articles 
