<edit: errr, that would be the outside of the platter that has a higher angular velocity?)>
Yes, the outside should be the fastest--the most surface area passes under the head at the same rate of rotation. (= same angular velocity, higher linear velocity.)
Here's one pretty bad case that might give you an example of what to expect: I have 120GB data-only drive that I use for a bunch of little files, plus several large DV projects as well as a collection of files in the 500MB - 1GB range. It has been run under 10.3 for almost a year with no care other than fsck on occasion, and for most of that time it's been well over 90% full, on several occasions with essentialy no space left.
This is close to a worst case scenereo--little free space, large files, long time. As it turns out, less than 0.7% of the 46,000 files on it are actually fragmented. Of those, about four files are badly fragmented, and another 10 or so aren't looking so hot. Of these, all are over 600MB in size, which isn't at all surprising.
The only time I've seen any actual indication of slowdown is when the disk is above 98% full and I copy a relatively large file to it, at which point there's quite a bit of disk activity and performance is noticably slower.
The point here being that OSX does a reasonably good job of keeping fragmentation to a minimum on its own. It's not perfect, and if you have a drive without a lot of free space that you're moving big chunks of data on and off of (video projects, mostly), then it will get fragmented.
For the majority of people, it just doesn't seem to be worth worrying about, though some people who do a lot of video will probably benefit from either a 3rd party defragger or periodically erasing the drive.