WD's have a pretty good reputation, there was a bad batch of 80gb and 120gb hardies - where we were getting about 30% defect rate on them, but they were fixed up pretty promptly. Haven't had a problem ever since. And i'm running a total of 8 WD's in my various workstations... 2x250gb, 1x74gb raptor, the rest are assortments of 120gbs and 80gbs.
Currently my overclocked A64 3500 w/ X800pro powering a 24" widescreen lcd is on 24x7 as I use it to Fold for Overclockers Australia folding team. CPU usage on it is always sitting on 100% because i donate the spare cpu cycles to Folding@home. The uptime on it currently is roughly 6 months straight... i basically never turn off the computer unless I have to. If it means, my computer will die prematurely in 8years time instead of 10 years time then so be it, it'll be a dinosaur by that time anyhow.
But the computers at work run 24x7 and never ever go down... the longest uptime i have is on our webserver (http://www.dmac.adelaide.edu.au ) which has an uptime of about 400 plus days. That stays up even when there is no power to the building, and even when one of our hard drives or power supplies blow in the computer. Computers have become so robust in the past decade or so, that reliability is no longer a factor.
Edit, having said that, the weakest links are always things like cooling systems (refridgerated/watercooling/fan systems) they're usually first to give way. But modern motherboards will often just shutdown before excessive heat is introduced preventing a meltdown.