You probably know how matching fuel injectors for a multi-cylinder engine can lead to a good performance increase. But what you probably not know is that not all fuel injectors are created equal, in fact a Kawasaki racing team took a bag of over 25 injectors to a well-respected firm that specializes in fuel injection and had the necessary equipment to measure the injectors properly.
The variation in flow rates was eight percent although I can't recall the actual flow rate, and so they matched the injectors to suit muliple-cylinder engines. The best set of injectors varied by less than one percent and the next best set wasn't much more – so in terms of matching them all together, they were reasonable close.
Race teams spent many hours on the dyno, mapping both engines and have very well developed maps for the engines. Essentially maps were the same but varied slightly as the motorcycles had different con-rods and picked up at different rates. On the engine with the best-matched injectors they found no measurable difference in the air/fuel ration (measured at both the mid-pipe and tail pipe) or power output. On the engine that got the meanest injectors, the air/fuel ratio was slightly leaner, but only 'points' of a ratio. A trim of one to two percent corrected this. While the exercise wasn't a waste of time, it suggested that unless you have a very bad set of injectors, or far more accurate methods of testing, there is little point in matching the injectors for most people. Of course there will be exceptions – but I would say only in advanced applications.
In terms of measuring the effects, one of the problems was looking at each cylinder separately. The Kawasaki race team used a four-channel lambda module with an additional probe watching the overall mixture, but the problem has always been supplying enough heater current. A typical lambda probe draws about two amps when heating, so there was a maximum draw of 10 amps in the system they where using. This may bot be a problem for a road bike generator but it's not much use with race generators. On the motorcycle they tested the road generator supplied 30 amps at 5,000rpm but the kit system could only supply 10 amps at 8,000rpm. So the battery went flat over a number of laps.