Well, I have to make two points.
First, the OP's OC is just awesome anyway you look at it. Whether it's the chip or his skill, hell, being able to POST @4.0GHz+ with a 2.8GHz/$280 chip is already miraculous to me. Can you imagine FX57 @4.0GHz? Even just a POST? I understand that OP claimed "Being stable" first, but we should all first congrat his (or Intel's?) achievement. Awesome chip/job, D94!
Second, I think by now we can somehow agree on "Stability". Yes, stability can be relative but then again so is everything else. If we decide to be "relative" on everything, discussion becomes impossible. That's why there are common sense and statistics. Say, if you're given a choice to pick a chip that does 12hrs Prime95 and a chip that does Super PI 32M, which one would you pick? With Prime95 unstable setup (especially LargeFFT) I have had quite a few mysterious lockups, reboots, etc. while playing games. Yes, Prime95 is not perfect but it's the best tool we can agree on for stability of a CPU. So at least for me, (and hopefully for many others) it's like this.
1. Prime95 SmallFFT: CPU Stability
2. Super PI: Memory Configuration
3. Prime95 LargeFFT: Overall System Stablity
Bleh.. this is not what I was going to say.
Anyway, congrats again on great OCs. It's not just the OP's achievement. But it could be a sign of not-so-distant future. Think about it. Socket 939 Opteron? Do you think AMD all of a sudden decided to lose money on their chips by selling those overpowering chips at fraction of the prices? I don't think so. It means AMD's manufacturing has leaped to a different level. It's also obvious from the fact that these chips are showing a very uniform performance pattern. On the same line, this Pentium D 9xx series might just mean that Intel has finally mastered the Netburst architecture and the manufacturing. Unfortunately for them, it's like the last minute spurt since a new architecture is coming after this. Still very impressive. Just like Socket 939 Opterons.
After all, we're observing a huge leap in CPU manufacturing that has been stagnant for such a long time. From both AMD and Intel. Arguing over stability is somewhat off the point, IMHO. If someone can POST @4.5GHz, the chip could be stabilized @4.0GHz by someone else. What's more important to us is the progress.
P.S. Mark, your posts are getting worse and worse.