1) Max power dissipation of pure 20 fan : 2.4 W ..
Even if efficiency sucks ... 2.4 Watts all together ...
2) Heat management comes first ,in order for the cobs to operate as efficiently as possible ,no matter the Io* .
A good cooled COB will operate at higher driving currents ,with same or even higher rad. efficiency ,
than a moderately cooled COB ,driven at lower current than the first one.
Light spread comes second regarding designing ,as instead of a huge long heatsink ,3x smaller ones could have been used in place ,each bearing two chips on it( and with it's own fan

..).
Or appropriate optics can be installed ,also ,to manage light spread.
Light spread issues can be solved by several / few alternatives.
Cooling issues definately share not the same possibility.
3) Nice thick base of .312" ,24" long ,1.312 high ,7" wide ....
View attachment 3323664
View attachment 3323665
As I said ...Even when 130 cfm of air (pure 20 ) are impinged ALL OVER the fin surface area ..
( A LID IS THEORETICALLY USED ) ..
Still the heatsink is too long ...
The 200 mm fan ,at max and with a lid on ,still will have trouble evenly cooling those cobs ..
But,that is merely ,only my opininion.
Some real life testing ,with a thermocouple attached to the appropriate Tc points of cobs ,
can reveal more closely to the truth ,what actually is going on .
I'm really hoping to be proved wrong about my assumptions ,on this topic .
Cheers.