Hmmm as a long time tinkerer with Raycal (new Spectra) I'm wondering what your final spectrum criteria is?
Couple of things to point out.
The ATI data is based on 39W tubes.
The led generic data is based on 1W each. Realistically one should use like 2W (2X) for normal arrays of "3W" class diodes.
You can use fractions such as 54/39 = 1.38 for a 4' tube
Blue plus 3x 1.38 = 4.15 rounded up to 4.2
I'm not seeing what you are picking puts much of a dent in the heavy blue of the tubes.
Second whites can add a tinge of yellow esp. warm whites though in a sense less bothersome than the cool whites
due to the low red/high yellow-green content.
Third every white adds more blue to fight against.
Ignoring the violet/uv for the moment.
Now that is on paper, visually is another story.
"Scientifically" all you want to do is add more green and more red to blend with the blue producing white.
Blue Plus is not full spectrum. You do know that right? Coral plus or aquablue special are closer to gull spectrum. Fair enough . Coral plus. Not even close to full spectrum daylight. Seems to be missing HUGE chunks of " colors". These are from ati btw. Since most people don' t like the...
www.reef2reef.com
Takes 300W of green and 100w of red to push your blue centric spectrum to a 90000k spectrum.
Better balance:
To be honest the wattages needed seems ridiculously high. Not sure why.
Point still holds, adding red/green/possibly cyan is arguably better than adding whites.
Turning a single 39 watt actinic tube to a "daylight" spectrum isn't terribly hard.