I'd stay away from using any benchmarks on a VPS. Benchmark software is designed to max out the hardware in a non-production environment. If you do that in a VPS, your host won't be happy.
Often, I find that a service like websitepulse.com can help. You can have it measure the response times of your web server from several locations. The one with the lower web server response times is the faster one.
Or, you could use the old standby. Which 'feels' slower?