Actually, assuming that two disks each have the same individual MTBF, then having two disks instead of one doubles the chance of failure. Nevertheless, when a single failure occurs, the redundancy makes the fault tolerance provided by the second disk infinitely better, so the system is...