The best answer I can come up with is ''not much''. Even if the rumors are correct, and the Mac Mini and iBook get Intel chips, there's not going to be any immediate need to start wholesale purchases, for a few reasons.
First, are you looking to replace machines in the January/Q1 2006 time frame? If you are, then you may be looking at these new machines. If not, then while you should be interested in them, they aren't critical.
Other than for test purposes, I wouldn't rush out and order a pallet-load of them anyway. You'll want to spend some time testing them in your environments, to make sure the software you have now that isn't Universal Binary runs properly for you under Rosetta. If you still need to run Classic for some applications, then you can't go to an Intel-based Mac, as the Classic environment is not supported, nor will it run on that hardware.
Second, if you're a heavy Adobe shop, or your users need high-end processor intensive applications that aren't yet native under Intel, you're really going to want to wait for native applications. Yes, Photoshop may run well enough under Rosetta, but do you really want to saddle someone with ''well enough''? Probably not.
In my case, we're looking at some Quad G5s for our designers. This way, we don't have to upgrade our software right away to support the hardware, (that's important since we still use Adobe CS1). Considering the costs involved in an upgrade to the entire CS Pro Suite, this is not a small consideration. When the Universal Binary version of CS comes out, we can upgrade to that separately from any hardware upgrade, then look at Intel-based Macs in our next hardware cycle. I don't see the Quad G5s suddenly becoming too slow to use anytime soon.
When you see new hardware, and it's shiny, and pretty, it's easy to forget that it's only one part of the equation. You have to have the software for it. You have to know what changes you'll need to make in your support and management procedures for the new hardware. If your mission-critical applications don't run natively, and you don't need new hardware right away, don't just upgrade because you can. New hardware like this is always a bit of a pain anyway. Don't subject yourself to it unecessarily.
Until the software my Mac users need is native under Intel, I have no problems ordering PPC hardware. We still get newer hardware, and Apple still gets the sales. I'm not putting off purchases en toto, just delaying Intel purchases a bit.
You'll also want to see about the host of other differences that are going to crop up. Will FireWire booting be supported? It should be, but until we have real hardware shipping, we don't know. What about PXE booting for Netboot? We don't know. What about improved Wake On Lan/Power On Lan? We don't know.
Since there's going to be a different version of OS X on the Intel Macs, (not necessarily a different number version), will it be Universal out of the gate, or will we have to wait for 10.4.4/.5/.6, etc? We don't know. What about dual-booting Windows? Will that be easy, painful, or in between? How will OS X deal with a local NTFS drive or partition in that case? How will XP deal with this? We don't know. In a dual-boot situation, can you PXE-boot for WIndows OS installs? We don't know.
There's a ton of minor questions that you're going to need a few weeks, maybe even a few months, to answer. ''Well, let me know how it works for you'' is not something you want to drop on a lot of users. So even if Apple does release Intel-based Macs, you're not going to want to start deploying them immediately. You'll want to test them first.
Of course, for my company, I suppose that means I'll have to bear that burden, and sacrifice hours and days playing with... er... I mean, testing the new Macs. But, I suppose I'll manage somehow.
Sometimes, being in IT doesn't suck.