I personally believe the SR view is a little overstated.
Of course STR dependent applications will benefit more when the hardware can be configured to improve STR. One thing I believe that SR lacks is a firm grip on the priorities of the 'RAID 0 happy' gaming public. According to SR's own research STR is over 25% for typical use (Office DriveMark 2002 - http://www.storagereview.com/articles/2001...issance_6.html). Of course this is clear evidence that that the majority of applications will not benefit from improvements in STR performance.
But I will tell you, if I look at RAID 0 and estimate, worst case, that I will get a 50% improvement in STR, then I look at SRâ€™s typical 25% of STR dependant applications and think "what would I pay to be 12% faster here? In fact, what would I pay to be 5% faster than the next guy?" One answer: 'As much as my budget allows'. High-end gamers are not skint; they simply don't worry about dropping a couple of hundred bucks to get small increases in general performance.
So again, to me at least, SRâ€™s view somewhat misses the mark. I know the tweakers.net article goes someway to remedy this, yet with only StorageMark 2004 as their best tool their tests are not ironclad. Either way, I have, spent my couple of hundred to gain my small percentage in performance, and I am a happy RAID 0 gamer