Xapplimatic: Unfortunately, your testing methodology for speed doesn't quite work. You can't test on a non-local server that you have no control over, because there are plenty of issues in terms of server load and fluctuations in connection speed to skew your tests. If you want to do a speed comparison, you should do two things:

1) Store a file on your hard drive and do a true test of page rendering times. This tests the speed of the renderer itself (no networks in there). Make sure nothing else is running in the background. Load the page several times. Take the average.

2) Store a file on an isolated server on your LAN. Make sure it's a wired connection (there are issues with interference in wireless connections that might cause problems). Have nothing else accessing the server. This way you can test how they handle packet downloading. Load the page several times and take the average.

I'm not refuting that OmniWeb is faster. It very well could be. But testing it on Apple.com (a dynamic homepage) on a dialup connection and being off by 2 seconds is pretty much explainable by simple fluctuations in your internet connection and server load.