just saying that the quality control on sata is 1 out of 12TB and that in 2009, the avergae disk space will increase to the point that the raid set up will eclipse that mark and thus fail (if raid set up is using large hd's)
what needs to happen is improve the error accountability