Not true. At least not according to Garrett. I've read some of the 3" vs 4" housing debates (not sure if I saw Jr's contribution as it was a couple years ago). From what I do remember, there was no direct comparison data available where the compressor housing was changed on the spot with the same features (surge ports, polish, no inlet/filter, etc.). What I have seen is LOTS of assuming '4" is bigger so it has to be better' statements, which might be true for tip-in response when you're driving an 8.9 liter Peterbilt tractor trailor, but means much less when considering ultimate performance.
The actual inlet for most of us bigger turbo people is in the 2.2-2.4" size range. Take a look at the inlet of any 4" turbo and the inlet is anything but a smooth transition. It's actually pretty freaking horrible! So why Chris did the manufacturer's (and performance industry) in their great wisdom move from 3" to 4"? In one word, surge. The surge ports and surge characteristics are greatly improved on the large housings. This provides significant advantages in tunability and power capability lower in the revs where a 3" non-ported housing would have to be tuned around the surge line leaving power potential on the table.
If my logic is making any sense here you can hopefully see where we're going here... Combining the best characteristics of both. A trumpeted 3" inlet with the 4" housing's surge ports. The theory made sense. I passed my designs to Garrett for input and according to software simulations the "Killer B Motorsport Compressor Housing Inlet Adapter" has improved surge capabilities even over the 4" inlet because the surge flow is much less turbulent. As well as marginal improvements in overall flow due to the smooth transition of oncoming air into the wheel.
Now I'm not going to make it sound like it was all planned out this way, because it wasn't. Like many discoveries, I came across it entirely by accident when I decided to reduce the 4" inlet to 3" due to the small clearances around the inlet. It was really rubbing me the wrong way, but I figured if I've got the ear of Garrett engineering they'd be able to guide my path and/or let me know if this would be a problem to the point where I'd pull the plug on this kit. I was pleasantly suprised with the feedback I was given
So there you have it... Or at least an understanding of what we did and why. I don't have back-to-back REAL dyno data to prove it, but I do have a kit on a car that makes more power, and sooner, than any kit I've seen with the same size turbo, fuel type, and comparable mods. The GTX3576 should definitely be interesting... better yet the TS GTX3576 should be even more so
I'll give you partial credit on this one since a reduction in size does make a 1/2" offset from the original size