Dan, I have said this before, and I'll say it again: I am not a fan of Prescott. It's ridiculous to consider what management idiocy could have resulted in a CPU with a smaller process, more cache, a longer pipeline, and a completely new redesign of the core could not manage to outperform its predecessor at launch. It's a complete failure relative to what Intel ought to produce from their desktop division.
On the other hand, with respect to this topic, all this is irrelevant. Assuming that Intel can convince end users to buy a Prescott instead of a Northwood - and for the most part, they will - then they will have the benefit of a smaller die with 300mm wafers, and the costs will be reduced. That has been my only point, and despite what you might think, Intel is already selling Prescott by the millions. Even if 3.4GHz parts are in low volume, these are at the end of the demand bell curve, where volumes are not important. Intel's highest volume processor is probably the 2.8GHz model, which is will within Prescott bin splits.
Intel's sheer manufacturing power allows them a benefit in this kind of market, but I don't think they can use this power against AMD without more competitive performance. Athlon 64 is getting the top bin ASPs, which could earn AMD a small fortune while it lasts. The only thing people here should be asking is how long will this golden goose live?
Market Data 
Markets 
AI
