Since you have already created a branch with your VDB's then the data is completely different and separate, no matter how identical the composite. . In a tree analogy, you might have identical twigs and leaves, but you are two different branches now.
In that model, and if the amount of data is that significant, I would suggest that you have a Trunk VDB and apply the large update to the trunk VDB and then refresh the branches (VDB_X & VDB_Y).