Yeah, good point, the magazine does have more resources than the rest of us for testing accurately...heck, I don't even have an
SG I can hook up to my car (OBDII was another 29 years away
), though, so observing the consistency of results is all I have, much like the way they conducted their test...though over a longer period of time than a mere 52 mile test course. 7 tanks in a row, immediately, over 17mpg (a 20% improvement in my normal 14mpg FE) was enough to tell me that leaning out my carb and backing my cruise speed down to 40mph was statistically a valid increase. If it'd been just one tank, or some 52 mile test loop, though, I wouldn't have been convinced at all and would be equally dubious had I been on their staff.
If they insist on testing that out without the dyno time, then I'd even tell them to average out a few weeks' worth of driving before and after and see if there's any kind of improvement that's statistically valid. Like you, I don't see how a K+N is going to account for a 10% FE increase.