Spray pattern affects effective mixture, through puddling, poor atomization, poor mixing. The effective mixture is what produces the quality of idle, acceleration and power. The O2-sensor reads the results of this effective mixture and trims the mixture richer. (The O2 sensor senses oxygen concentration, not A/F ratio or unburned fuel.)
Point is, spray pattern affects the quality of the air/fuel mixture, and the system is likely to react by increasing fuel flow to compensate.
The key to my quote is 'under low vacuum'. As in vacuum (airflow) inside the intake in response to throttle opening. At cruising speeds' ie: 2k and up, the puddling effect then does one of two things depending on the intake design. It either atomizes poorly, or turns into a micro-stream and enters the combustion chamber. This, in fact, causes the O2 sensor to see it as rich since the fuel does not completely combust and leaves very little O2 behind and the ECU will trim the mixture leaner. And the O2 DOES measure AFR but based on the O2 concentration left after combustion. In perfect combustion of s 14.7:1 mixture, there is a specific amount of O2 left based on the volume of the engine, the volume of the air in the combustion chamber and the amount of fuel injected. The ECU has a predetermined 'map' that knows what this amount is under many different operating conditions. If the O2 goes over or falls blow the map value, which is an AFR of 14.7:1, then the ECU adjusts accordingly. Standard 3&4 wire O2 sensors are no where near as accurate or sensitive as Lambda O2 sensors, but that is what they do in a round about way. Not trying to start a pissing contest, but I've been a licensed emissions tech for 10 years, built custom intakes, and have done fuel mapping for custom injection systems, so I do have some insight here. The point being, cleaning injectors has little to no effect on mileage in the majority of vehicles on the road. There are of course, always exceptions.
