High 2D band can be due to overlapping single layers. Turbostratic graphene. The layers are not electronically coupled and behave like a sum of single layers.
The G band represents the disorder factor in your graphene. The edges of the flakes are part of the disorder. So if the beam of your Raman spectrometer is covering much smaller flakes you wil measure potentially a lot of edges. Than the final specter will eventually not be representative of the quality of your graphene or GO by introducting disorder where there is no reason to be considering your cvd technique can make very good graphene.