The October issue of BMWCCA's Roundel Magazine has a good article on Dyno-testing cars entitled 'Numbers Don't Lie'.
Evidently, attendees at the club's Oktoberfest in Indianapolis had the opportunity to verify all those claims about how much horsepower each actually was putting to the street.
According to Kenny Brown, who made his equipment available for the tests, drive-train losses typically run between 14 and 20 percent on all cars. Thus, the 305-hp Mustangs that make up a significant client base for him --- see between 250 and 260-hp at the rear wheels.
One key point in the article is that people often (incorrectly) assume that one can simply add up all those claimed HP gains from chips, exhaust systems, free-flow intake, etc. to get a total expected hp figure. He points out that in the real world, the claims often fall short --- and certain vendors hide behind words like 'as much as 9 hp', when typical results might be much less.
Unfortunately, there was no mention of stock E46 tests to compare, for example, actual at the wheel output of the 323 vs. the 328. Much has been mentioned here about supposed 'under-rating' of the 323, ---- which to be honest --- I find unlikely. OC, only collection and posting of actual data will tell one way or the other.
Perhaps, it's time for Ausgang to call BMWNA again, and see if they would be willing to provide any data. After all, dyno tests are routinely performed at the VPC centers. I've watched them done at the NJ VPC. Maybe I can get some figures for median hp on 323s and 328s, and some measure of variability, (ideally, the standard deviation). Without the variability data, it would be easy to make the mistake of generalizing from only one 323 and one 328, which may not be 'representative' of what most of us are driving.
Will post any data that BMWNA provides as not proprietary, and if I have any luck reaching the right person.