Regarding the SEW study, here's some commentary and critique on WebmasterWorld, from back in July....
Has the Anchor Text Signal Been Dampened? http://www.webmasterworld.com/link_development/4175042.htm [webmasterworld.com]
IMO, while the study is interesting, its approach is flawed because it considers inbound linking in isolation. I've always felt that inbound anchor text, link quality, page titles, onsite navigation, and onpage content all work together.
With regard to the domain name advantage... in areas I observe regularly, there are many exact match domains that are apparently unbeatable, and there are some that aren't ranking well at all. It's not always the .com TLD that's the winner. Obviously, what you do with the site has something to do with how much mileage the domain name will get you.
Often, the top exact match domains in very competitive areas are also extremely good sites, and they benefit from years of success and development dollars. Keyword domains that don't do well are generally cookie-cutter sites hoping to leverage the domain name by itself, with not enough emphasis on development of site content.
I do feel that keywords in the domain... or really
in the company name... give a site a great anchor text advantage at the start. This is such an advantage that in some areas I'd say it's extremely hard, if not practically impossible, to beat established keyword domains. I think that it is the anchor text, not just the domain name, that carries the weight. More precisely, it's the inbound anchor text in combination with keywords in the domain or company name that appear also on the site.
Google's addition of algo branding factors was perhaps intended to help level this playing field to a degree, but the increased weight given to brands has also served further to push newly emerging sites down.