I'm not even going to go into how many levels you're just plain wrong on theories of airflow.
And just so typically you don't even try, you just assume, this is why this site is widely regarded as one of the poorest tech sites for real information.
Why don't you try being less arrogant and try such cases tests your self? YOU WILL find like we have from our heat tests that it does make a SIGNIFICANT difference.
Just because you didn't read it on Anandtech first doesn't mean it isn't true. Go out and find something out for your self for once.
OK, I'll bite.
I can't believe this, after all the marketing that companies like Silverstone have done to inform people about positive air flow you still do you review in the completely silly and unrealistic non cased on desktop testing.
If you'd have read the original 480GTX review like I linked, you'd know that we tested the 480GTX in a Corsair 800D. With more intakes than exhausts the case has positive pressure, and guess what, thermal performance was still shocking in that case. If a $300 case failed to keep the 480GTX card cool (when every other card used in it coped fine) the 480GTX just shouldn't be that hot to begin with.
It has been pretty obvious that ATI deliberately created its newer cards like the 5970 to be so big that it forces the user to use a bigger positive style air flow case.
Really? This is the reason they had to use a 12.2" PCB to fit two graphics cores, 2GB memory, the bridge chip and all the power hardware onto a single board? Yep, Definitely obvious.
If you have a proper positive air flow case it is like the card it self has a full 20cm fan attached to it instead of the tiny fan it does have. Let the case do some of the work in cooling, new positive air flow cases from Silverstone can aid in cooling the card down not negate it. Or try the CM Storm Sniper with 2 x 20cm fans on its filtered side intake area, that should give some good positive air flow.
It doesn't work like the card has a 20cm fan attached to the side of it. Most of the positive pressure is going to escape out of the path of least resistance, through a densely packed heatsink graphics card exhaust is not that path. Sure it'll help a bit, but if you're selling a product like nvidia is, you can't rely on the end user having a certain case setup for your card to perform thermally well. What happens when you start to run these cards in SLI configurations? Positive case pressure isn't going to do much for the cards then.
Cases I've seen with huge fans on the side are cheap (or cheap looking) cases, and if someone is spending $500 on a graphics card, they'll at least have the money (if not the sense of style) to buy something more than a tacky looking case with a huge 200mm+ fan on the side.
A video card like this should be a case manufactures dream and we don't have any reviews out there showing off the potential yet.
How is it going to be a case manufacturers dream to design a case around cooling a single card, that is ridiculously hot, that is going to rack up barely single digit percentages of the total cards nvidia sells? Lots of return on the R&D investments there.
Bottom line though; you're completely attacking the wrong article for what in your eyes is bad review practice case setups. This particular piece is meant to be a tongue-in-cheek look at the new nvidia vBIOS (designed to work out issues with multi monitor support) and a play on the old adage of "it's so hot you can cook an egg on it".