Some say that PWM fans are design to operate at 12V always and they regulate internally, some say lower voltage might be risky for its design... but Noctua PWM fans even come with a LNA (a resistor) that lowers the voltage more and changes their operation. I try when planning my setup to run 3pin on voltage contolled or undervolting with fan controlers, and use 4pin pwm fans on CPU_FAN headers that manage them via PWM signal.
One thing worth mentioning is that in the last two weeks i have tested mulitple PWM fans on both voltage controlled headers and on pwm headers, and at least with FanXpert2 and noctua PWM fans on voltage controlled headers sometimes it does the test and stablishes minimium rpms but are not sustainable, the fan stop after a minute or two, like it calculates it to low and takes it where the fan spins down, using the noctua pwm fans on pwm headers never i have seen this or when using 3pin on voltage controled headers also not an issue, but this shouldn't matter much for someone looking to do it manually with a fanmate2.
On a side note, I'm not particularly impressed with this NF-F12. I saw a review on overclock.net stating that the fan is "almost completely inaudible at 1200rpm unless you have your head within 2 feet of it". Well at 1200rpm I can easily hear it across the room, even during daytime with the increased ambient noise and a window ajar, AND my loud computer running. In fact, as I write this it is raining lightly on the tin roof, and yet I can still
faintly hear the fan. I don't think the main body of noise is the motor or bearings, I think its some interaction of the air between the blades and the stators, although there is a a fair amount of vibration when I put it on a wooden bench or something.
Another interesting thing is that if I listen really closely when I'm turning the fan off, I can hear a 'tick....ticktick' sound as it spins down - just three faint taps that sound like hard drive seek clicks. And when I'm running it at max RPM, it only 'ticks' once.