FPI isn't everything either. Take a look at this radiator covered earlier this year:
https://www.techpowerup.com/reviews/HardwareLabs/Black_Ice_Nemesis_GTR_360/3.html, it has 16 FPI fins in a 54 mm thickness so if you look at those two numbers one would expect similar airflow restriction to the Black Ice Nemesis GTX but they end up on opposite ends of the spectrum with the Nemesis GTX a low-airflow optimized radiator and the Nemesis GTR a high-airflow optimized radiator with higher airflow restriction. This is why I chose a single radiator, irrespective of thickness, FPI, number of rows etc, with an average airflow restriction to help give a better idea of how these fans would fare for watercooling in general.
Not suggesting it is ... it kinda goes without saying that when discussing a factor in perfomance "all other things being equal". A thicker rad has more resistance then a thinner one, fin design fpi, materials, etc all being equal. But when making selections, criteria must fall within certain ranges... don't expect a 6,000 pound vehicle to win any mpg awards,
I own 3 of these fans on my EK PE360 rad and I love them. First time having all PWM fans in my system and its fantastic.
PWM fans are no longer required to provide PWM like control. PWM control has ... or should I say "had" clear benefits. The upside of PWM control was greater low speed control because even at slower speeds, they still provide a 12v signal to overcome inertia of the blade assembly. Was a time where only PWM fans could do this. On the downside, a) PWM suffered from low speed clicking or hum, b) cost, c) on some brands (i.e. Corsair) you couldn't gang more than 5 or so on a single channel and d) MoBos had a limited number of PWM headers. problem is, those old posts are still being re-quoted without taking into account technology changes that have come into play since then and d) has been eliminated since Z97.... was supposed to be Z87 but even tho manuals typically said everything was PWM, MoBo manufacturers were unable to deliver that to market.
a) DCV controlled fans now are capable of much lower speeds due to changes in MoBo design which allow for ramp up and ramp down of fan speeds. In the past, when system called for low speed and hence low voltages, the signal was not strong enough voltage wise to overcome inertia and get the blade assembly moving. Now, the signal will initiate movement with a higher voltage and then slow it down thereby allowing DCV fan speeds near 20%.
b) The advent of PWM Fan PCBs which take a PWM feed from MoBo and convert the signal to DCV with the same capabilities as above.
Which way to go ? Kinda a 6 of one, half dozen of the other thing.
a) Noise - as long as ya avoid the clicky / humming PWM fan models, won't matter significantly
b) No. of fans per channel - Again, as long as ya avoid the fan numbers or models known to fail here, doesn't matter
c) Cost - depends on number of fans.... using the example from previous post, DCV are $10 / PWM $17
3 DCVs with a PWM PCB would be 3 x $10 + $20 = $50
3 PWMs would be 3 $17 = $51
So less favors PWM, more favors DCV.
In your case with 3 x 120mm, Id prolly go with PWM fans
In my case, DCV makes more sense ... I have two rads (140mm x 3 and 140mm x 2) in push / pull plus 6 case fans and MoBo has 2 CPU and 3 CHA headers.
CPU => Pump No. 1
CPU_OPT => Pump No. 2
CHA_1 = > PCB No. 1 => (6) 1,250 rpm fans in push / pull on 420mm rad
CHA_2 = > PCB No. 2 => (4) 1,250 rpm fans in push / pull on 280mm rad
CHA_3= > PCB No. 3 => (6) 1,250 rpm case fans
Fans range from 325 rpm to 550 rpm typical (gaming) ... under stress testing (RoG Real Bench + Furmark) get to 850 rpm where fans are not audible.
(5) DCV fans came with the case so ....
Cost of buying (11) more + (3) controllers was $170
Cost of buying (16) PWM fans was $272 not counting (3) ModMyToys (or Swiftech, whatever) Hubs
Hard to go wrong either way, both options deliver comparable results