Mine arrived today so I did a little test:



As you can see there is little little difference between the two brackets. Can this "hi-flow" bracket really reduce temps 5C? We'll see.

SETUP
Card: eVGA GeForce GTX 480 SuperClocked (Stock speeds: core 725, memory 3800, shader 1450)
Driver: 257.21
Rest of the system: Check out my sig

Here's a picture of the system layout:
Please note that there is a physX card (9800 gtx+) installed right next to the GTX 480 so my results maybe little higher than most of you.



While I am at it, I also decided to check my good ol' 9800gtx+ which now serves as a physX card in my system.



Waddaya know! It didn't have a hi-flow bracket either (Are those wood screws I am seeing?). So, to test whether eVGA bracket bogus or the real deal, I have decided to create my own hi-flow bracket for the 9800gtx+. If this works, it should work for both cards (right?).

10 minutes quality time with my dremel and I got a home-made even higher-flow bracket.



TESTING METHODOLOGY
Average room temp 75F (23C)
Idle temps were measured after the system idled for 30 mins (temperatures settled for at least 10 minutes).
Load temps measured after a single 3dmark06 run. Temps taken right after the 3dmark06 run completed.
Temps were measured with eVGA Precision.
(I know my testing approach is very crude but we're measuring a relative performance here)

RESULTS

*Drumroll

GTX 480:
Low-flow: idle: 57c - load: 82c
Hi-flow: idle: 55c - load: 82c

9800 GTX+:
Low-flow: idle:63c - load*: 65c (*not actually load since it wasn't being utilized)
Higher-flow: idle: 62c - load: 65c

CONCLUSION

1- It's bogus. The so-called Low-flow bracket will not give you 5c in most cases.
2- Save yourself 5 bucks. Do-it-yourself bracket will not cost you anything. If you have the right tools, the process is easy and will not involve any risk to your card.
3- Even if you take the bracket off you won't get any difference in temps.

So there you have it.