You mentioned Frequency Balance in the first post and i think this is an important issue we need to get back to and discuss. Two identical fans spinning at the same RPM with have roughly the same spectral balance of sound if all else is equal, vent openings, air impedance (loading) etc. and acoustically will follow the +3dB rule.
However. several other factors also come into play. one is the point source rule. no two objects can share the same point in space and when you have two noise generating devices you will also have additive and subtractive acoustic effects due to the distance between them. this is called comb filtering. Without getting mega technical (If you wish, start here http://en.wikipedia.org/wiki/Comb_filter
So in our example. we have two fans and lets say they are spinning the same speed. depending on where the microphone is in relation to the two fans. the spectral content will change. if you "pan" the microphone across a plane from the fans, you will see certain areas of the spectrum change as you move between peaks and nulls of a particular frequency.
Fans generate an almost white noise and as such this is harder to measure then it sounds in free air. But any monotone steady state noise such as harmonics, hum or noise created from stationary fins on the fans case, vent holes etc. that the fan creates can have effects beyond whats expected.
But vary the speed of one fan slightly and you can spread those peaks and nulls out over a larger area so that less are bunched up all at the same areas. just make sure the fan speed is not a harmonic multiple of the first fan speed. IE 1000rpm and 500 rpm. you would be better off with 1000rpm and say 850rpm or 625rpm etc. and by varying one fan speed you can find what works best.
or if you need to keep the total CFM the same, raise one fan speed and lower the other. if the fans are side by side you can also "direct" air flow slightly this way.