I hope to make sure that someone at Codeweavers sees this.
I tested the new browser Iron (Google Chrome with the google spyware removed by SRWare). I tested it on the Sunspider Javascript benchmark, which was created by webkit (which gets the best numbers on it). This measures only javascript performance, which is perhaps an important number for browser performance, but hardly the only or even the most important number (I even turn javascript off a lot of the time). Both tests were on my desktop, 2GHz Athlon 64 with 1G RAM.
http://www2.webkit.org/perf/sunspider-0.9/sunspider.html
For the first test I booted Windows XP SP3 and downloaded the Iron installer, installed it, and ran the test. The cumulative result total was 2618.4 ms +/- 3.0%.
For the second test, I rebooted the computer into Puppy 4.00 and installed the very same Iron installer into the special wine configuration that Codewavers had created for their Crossover Chromium (since Iron is essentially Chrome). I had to play around a bit to get Iron to run, but when I got it to run, I ran the same Sunspider Javascript benchmark, and the cumulative total time was 2237.6 ms +/- 1.2%.
Results: The very same Windows program running in wine on Puppy was 15% faster than it was on Windows XP3, running on the very same computer. I guess to be most accurate I should say that the very same program reported that it was 15% faster running in wine than on windows.
So I guess the people who say Wine does not impose a performance penalty are right 😉