since gigaflops are heavily reliant on Altivec optimisations for maximum ratings, the best comparison of a P4 i guess would be from a program that was optimised specifically for the P4.
but that doesn't provide a good real-world measurement. if you want to see how the two compare, then you'd probably be best to try out a program that you use on a P4, then go to an Apple store and try the same program on a PowerMac G4. it's hard to compare the two, but that's probably the best way.