In my opinion, what they are announcing is not entirely true.
First of all they mention bandwith problems as a cause for the delay. This just not seems right. Why? Well, because Apple doesn't use their own servers for file serving, they use Akamai instead, which has plenty of bandwith capacity, approximately 2000 Gbit, and uses servers all around the world.
So bandwith shouldn't be much of a problem, I suppose.
For example they served the Windows Vista Beta's, both 32-bit and 64-bit versions which were 2 or 3 GB's each. That didn't result in any bandwith problems as far as I know. Why should an update of approximately 500 MB be such a great deal?
Therefore, 500 MB is pretty much for an update. Normally the size of an update varies between architectures (Intel, PowerPC) and the testbuilds contain a bunch of debugging code. If the normal update will be around 500 MB, the combo and delta updates have to be much bigger instead, they're always.
Second, I think it's pretty strange they announce a testbuild (9C23) which was "released" 6 or 7 days ago and then claim they know the update will be further expanded and delayed.
At the moment, as far as we know, no "known issues" are present and no new OS X client seed has been released. They actually released a OS X Server seed (9C25) yesterday of which are no known details.
This could be reason enough to assume they're pretty near to a release. Usually they release both Server and Client on the same day, so when they release a Server seed, they have to postpone the client as well.
I encourage Apple to take their time for this update, but I think a 10.5.2 release is pretty near. I would say Friday afternoon (EST). I think they'll release a couple of normal software updates tomorrow to be ready for the big release the next day.
Of course, I could be wrong but I just not think the source is right about this.