I'd say that long seems excessive, considering that Safari loads much faster than that on my G3, which isn't even supposed to be able to run Tiger and also only has 640 MB of Ram, compared to the 1 GB that the iMac in question here has. Loading times for a program (which is what the bounces measure, though actual time would be a better measure if the bouncing does stop after a certain amount of time regardless of how far the program has loaded) would seem to be a reasonable benchmark, although non-scientific - as long as the exact same program and the same OS then it does provide some comparison of the performance of different machines.
_________________
Main system: custom C2D 2.66GHz tower/4GB RAM/2600XT 512 MB/500GB HD/XP Pro x86+Vista Business x64
Main Mac: Beige G3 minitower (Sonnet G4 500MHz/640 MB DVD/USB/Radeon 32 MB/OS 9.2.2+10.4.11)