It is pretty common for a 2" diameter crossover pipe/mid pipe to support ~1000hp. You have to consider volume and velocity against the goals/application.
Personally I am building my turbo LS hotside with 2.5" pipes because it is just cheaper/easier based on what is on the market in stainless j-bends/reducers; not to mention goofy long tube upswept Chinese headers. The mid-pipes are going to merge into a single 3" pipe into a T4 flange, and be vented by a 60mm waste gate. I am running a 4" downpipe off the turbine housing which will connect to a single 3" exhaust (hopefully made out of aluminum), and will have a 4" dump. On the intake side I am not running an intercooler as I plan to point the compressor right at the throttle body with a short length of 4" pipe and a 50mm BOV in between. Obviously I am running it all on E85. I figure adding a pressure transducer/MAP to measure the backpressure is pretty straightforwar d - could even run the math so it auto calculates the BP ratio in the scanner.
So while I am making a few compromises here and there, the ultimate goal is to minimize that variables that will influence spool time, hot and cold side restrictions, and backpressure. Plus, lighter is better. I've always wanted to go sans intercooler since I saw Bruce Plecan do it over a decade ago.
At the end of the day this "kit" should support a 4.8L engine as well as 454 stroker monster (which I will never be able to afford). But 4.8s, 5.3s, 5.7s, 6.0s, and 6.2s from the yard should be plenty for my needs/goals for years to come.
Essentially, any restriction you remove will reduce your "boost" - which is just a measure of restriction anyway and is going to introduce more heat into the intake charge as it works against itself. IMO, the best way to measure power is through mass air flow - it becomes pretty evident when your boost goes down but your calculated air flow goes up. When running a MAF it is a more linear and logical air flow model to follow than a VE table. My TBSS has a functioning MAF on it even though it is Speed Density tuned vehicle for this very reason. Less heat is more better too.
I've recently considered "building" a VE table using my MAF. Essentially I just need to log the MAF flow Hz against RPM and kpa and it should fall into place pretty easy. I am interested to see what the predicted model of the VE (uses baro, IAT, CTS and some other crap IIRC) versus the measured air flow of the MAF, and if they overlay equally.
GM uses a blended or hybrid Operating System (OS) where it will validate itself using the MAF and VE below an engine speed threshold (say 4000rpm) for all fueling calculations (and will default to the VE table if the MAF fails), and will go "MAF-only" above that RPM threshold for all remaining fuel calculations. Damn shame it isn't a Closed Loop OS.
