Furthermore, an anonymous benchmark (like many other anonymous product comparisons) is going to make many people suspect that it's a fraudulent result. Imagine there's an anonymous benchmark that isn't simple to replicate because of software and/or hardware costs that shows some proprietary software whipping an open source project on some dimensions or other. How many people here are going to immediately jump to the conclusion that it's the proprietary vendor planting false information?
Either way I wouldn't expect a real world benchmark to be reproducible, it's too hard to reproduce and takes too much resources.