As I figure it, you leave out the 50 IU number to get the formula
Correct. The scale shift number (50) is a design trick to get rid of the minus numbers in the entry IU argument and has no other application. For the individual trimlines, use the equation less this number.
Main tanks 650.7 as against 650.66
Centre tank 600.4 as against 600.14
Full tanks 628.8 as against 628.56
I'm surprised to see such close alignment. I would not have been surprised to see a bigger delta given the apparent round off for the IU .. surprises, surprises ....
I calculate CGs varying from 610 (empty) inches moving back to 650.7 inches (full main tanks) and then moving forward to 628.8 (full tanks)
Not hard to see some potentially big errors creeping in if one used a constant CG for the fuel calculation. This certainly can be done (and is quite common in trimsheets for smaller aircraft with non-prismatic tanks) but its use must go hand in hand with some appropriate error containment method .. typically an adjustment to the envelope limits as plotted on the trimsheet.