You really should use a test that focuses both on pure JavaScript and also DOM/JavaScript interaction. The fact that Chrome processes JavaScript faster doesn't necessarily mean that Firefox is slower in a real application that interacts with the DOM (although it may be, I'm not sure).

Edit: See http://dromaeo.com/

Sunspider and V8 are standard javascript benchmarks. I have found that the test results compare favourably to real world usage. For example Google wave is sluggish on firefox while chrome has no problems with it.