Microsoft FSX as a CPU Benchmark

I just read the article "i5, i7, Crossfire & SLI: Gaming Paradise, Redux" (excellent read BTW) and was miffed by the author's apparent dismissal of FSX as a good benchmark. I know it's not as popular as the other games benched. But I dare say it can put more of a load on the CPU than any of the others tested if set up like someone who's into flight simming would. Perhaps THG has some rule about using the software as it is out of the box.

Almost no one on the flight sim boards is flying the default scenary with no addons. Using the 3d cockpit, add-ons like Traffic X and some 3rd party aircraft and scenery greatly increase the demand on the CPU. This is a more likely scenario for most still into FSX. I have a QX9650 w/ SLI cards and can no way get playable frame rates at the settings they used.

Also, the review was to show performance capability of the various platforms and leaving the frame rate limiter on eliminates the value of the test for FSX. All the other games reviewed would be quite playable at ~30 fps too, but still they saw the value for comparative reasons of showing the 60+ fps scores.

While still playable at low frame rates, I have no doubt that FSX can put more of a strain on the CPU than any other games tested and hence still would add great value to the testing line up.

What do you think?

Exactly my point. So why not keep it as a CPU benchmark? I think a game's CPU demand, not it's popularity should be the mark of a good CPU benchmark.

It wasn't dismissed as a bad benchmark. The problem was that with the limiter *off*, results were so inconsistent as to be meaningless from one run to the next. Thus, it was left with the limiter on--as the game would be played.
One reader mentioned a certain config that'd help minimize the variance from one run to the next with the cap off. If you have any suggestions here, please feel free to let me know at cangelini at bestofmedia.com.