[Kevin M]

The coupling of thermodynamics and aerodynamics in an unclosed system is arguably the most difficult physics we know of above the quantum level. My point is that if someone without a pHD tried to explain why exhaust diameter downstream matters less than upstream, they're doing some combination of guessing, repeating something they read on the internet, or talking out of his or her ass. So no, I can't explain the phenomenon to you in precise detail. I can say with certainty thanks to more than a decade of relatively good empirical dyno results that, given good component design (meaning good flowing mufflers, if present) that a full 3" turboback makes the most power, a full 2.5" the least, and a 3" front, 2.5" back setup is in between, but favoring the full 3" performance. The reason it's not a 50% difference is that downpipe inlet design is a more important factor than diameter.

Sorry for pulling the de rigeur SECCS nanoargument on you. It's political season, which gets me all riled up about anything a person asserts when it's using false arguments, bad logic, hearsay, or "pretty well accepted" conventional wisdom. Yes, you are correct that a 2.5" aftermarket catback or axleback is a pretty minimal sacrifice in peak power on a stock turbo Subaru. But it is the claim that the colder exhaust needs less volume, which dates back to a guy from Turbo XS incorrectly applying the Ideal Gas Law to exhaust flow that annoys me. Largely that's attributable to me being an employee of Shiv Pathak at the time it happened, and people ignored Shiv correctly pointing out that the ideal gas law only aplies to closed systems like baloons filled with air and pressure cookers and the like, because he happened to be a huge douche about arguing with anything his competitors said on the internet, especially if it was wrong.

If you want my poorly educated guess, I would say it's just that the pipe diameter in the latter half of the turboback has become a second order factor or below. Turbulence, bend design, muffler baffling and design, all these are more important. The thing to remember about exhausts is that the single most restrictive part of it is nearly all of the power loss you get from that exhaust compared to having none at all. So, if the muffler baffling design is the most restrictive part of the catback, and it has a randomly assigned restriction value of 12, and 3" piping has a value of 6, and 2.5" has a value of 8, the total restriction is going to be close to 12 if you ignore other factors like bend shape and the transition from 3" to 2.5" regardless of which diameter you are using for the catback.