You can show a 1.0GHz XP beating a 2.0GHz P4 in benchmarks with the right benchmark. Likewise a 1.3GHz P4 can beat a 2000+ XP in the right benchmarks. Its not easy to compare when they don't use fair or optimized benchmarks.
btw - They ought to just release them as XP####A+'s, to denote the alternate core. This way they can call the larger cache T-bred something completely different. I'd like to see the XP####A+'s at a 166fsb, but its not going to happen.