Those are some interesting counter-points. I admit it's always easier to criticize something than to find points to defend it with, so I enjoyed reading your article Brett.
I'd like to make a few counter-points of my own however!
When a company shuns a website for not playing along and sends other sites review samples, then it only affects that single site. Not the entire industry, which is what occurred here. Readers of the site in question are free to decide of it what they will. If a company plays hardball with Anandtech or Tech Report then it's probably a company whose products I wouldn't want to use. The point is there is still free market in play here. And vice-versa, if the website did something a company thought was unfair, I'm still free to determine that myself and if so I could decide to change to reading a competing website.
In this situation AMD found a way to game the system. If the site publishes only half the story, AMD gets to manipulate first-impressions and more importantly capitalize on the early readership that won't be coming back to read the second half of the story. (I liked how a few sites labeled it as a full review to capitalize on this)
If the site does not publish the review, then they get penalized by a loss in readership (aka revenue) and secondly their competition will see an increase in readership. They first lose readers from the launch-day article, but they will also receive fewer readers when they do eventually release a full article at a future date. It's only anyone's guess as to how many of those same readers would make their switch to a different site a permanent change, but for readership that only reads tech sites for launch day articles, I would have to wonder.
As for my opinion of Trinity... (completely my opinion and not that of Techgage)
...for the mobile market the GPU performance is needed. But for desktops I do feel it has missed its niche. It's still not strong enough to make it suitable for gaming at 1080p (it's a desktop, gamers most likely are already buying a midrange GPU or may have one laying around that performs better). Users won't be buying it for its CPU performance either (the Core i3 3220 costs the same with 12% higher single-threaded performance, which the majority of programs perform best with). Finally, the CPU power consumption under load is 10 watts under double that of the i3 3220. This is important because it limits the chip from a portion of the very same HTPC and SFF market it is targeting. Again it depends on the type of user, since there's a wide diversity in the HTPC market. Some load leave a sustained load on the CPU and use it their HTPC as a general purpose server, others use it for video transcoding, some for gaming. Others leave the CPU completely idle.
The key point here is how we personally view the gamer/gaming market, and the HTPC/SFF users/market, as those views will decide how we feel Trinity meets the needs of these markets.
If gaming at 720p resolution with medium graphics detail and no AA/AF is gaming, then Trinity excels and is definitely a better buy than any Intel IGP. But if someone was actually buying this chip for gaming use, I feel it's more likely they would be gaming at the resolution of a desktop monitor (1080p) and would want settings other than "low detail" @ 30 FPS.