Look at it this way. For MAXIUM air flow at max speed and noise, with the 120's you get 3 x 41.5 = 124.5 CFM at 1700 RPM generating 20.6 dB noise PER Fan. with 140's you get 2 x 56 = 112 CFM at 1500 RPM generating 23.3 dB PER FAN again.
Now, what an automatic fan speed control does in fact is TEMPERATURE control - in this case, of the temperature as measured by a sensor on the mobo. It manipulates fan speed to achieve a TEMPERATURE target at that sensor. So if we compare those two fan set options, for any measured temperature the controls will slow down all the fans to whatever it takes to reach the target, and generally this will be to a speed that is a % of max speed. IF given the SAME % max speed signal, the 120's always will be going slightly faster than the 140"s. BUT they also will push MORE cooling air that way. So in fact the automatic system will see the temp getting lower, and let the 120's slow down more until the air flow rate they generate is very similar to what the 140's would do. BUT for SAME air flow, the pair of 140's would have to run at a higher % of Full setting. So their noise generated will be closer to their max noise figure.
There really is NO way to predict with math the actual speeds and noise generation for these fans base solely on their published specs. We can only make a qualitative guess. My guess is that, in actual use at less than full workload and max heat generations, the three 120's may be slightly quieter than a pair of 140's because the trio will run slower. That is my GUESS. I also expect that the difference in noise will be small.
Between these two options there is one other difference that MAY be significant for you. IF you expect to be running very high workloads and need maximum air flow for cooling, at top speeds the three 120's can move more air than two 140's. For workloads medium-high and less, either system will do a good job of cooling.