martinibuster - 1:18 pm on Jun 6, 2013 (gmt 0)
Site architecture plays a role in how a site is ranked because it plays a role in how the algos understand what content is important and from there the micro and macro meanings of pages, sections and the overall site.
I know of a recent example (of someone who came to me for help) where inbound link count changed so that an inner section gained more links than the home page and other sections. The home page stopped ranking for certain phrases but the one inner section picked up many more phrases. The inner section then stopped linking back to the home page section and the entire site, in the eyes of the algo, became known as being about the inner section, not the overall topic linked from the home page. Now when you search for the domain name, you get results from the inner section. This was caused by the abundance of inbound links to the inner section and was happening for awhile because of the natural inbound linking patterns. The ranking trend was pushed over the top by changes in the site architecture.
That and many other examples I have seen and experienced provide support to me for my belief that site architecture plays a role in what a site ranks for. Inbound link trends also play a role.
Given the above, I believe that if the top ten for a given niche have a particular linking pattern, such as an abundance of links to the home page, that that is a coincidence in that particular niche. It could also indicate an exploitable entry point.