It's about the subject area. Fowler is an expert on OO architecture, not low level optimization. I think you're holding him to the wrong standard, and I think it's rather absurd to insist that anyone has a duty to be what you think they should be professionally.
The way I see it, they're linked. The memory hierarchy has a direct impact on a number of architectural concerns: data organisation, choice of components, communication patterns between components, etc. It can make the difference between a system that meets its functional requirements and one that doesn't. These are all things an architect is supposed to care about, so to me it seems reasonable to expect an expert architect to be aware of the factors influencing them. I get the feeling we're not going to agree about this though.
It's about the subject area. Fowler is an expert on OO architecture, not low level optimization. I think you're holding him to the wrong standard, and I think it's rather absurd to insist that anyone has a duty to be what you think they should be professionally.