Well, the existence of the G mode means you have sp2 carbon in some form (graphite, CNT, fullerene) while the D mode is symmetry-forbidden for a clean structure and only is allowed if there are defects. That means your modification changed the nature of the defects in a manner that the allowance of the D mode vanished again; this can either mean you healed the defects (aka the perfect world scenario) or you modified them in a manner that the carbon around them is so far from sp2 that the D vibration is no longer characteristic of the material.
In order to differentiate, you could make a statement about the G mode intensity. Is it rising or decreasing?
Juergen has very eloquently summarised the G & D Raman bands in graphene-like materials. it is hard to provide more insight without knowing something about your GO modification or some idea of what your Raman spectra look like, but an absence of a D band implies a perfect sp2 hexagonal lattice. Sometimes these D & G peaks can be quite broad - are you sure that you have not ‘lost’ your D band by simply having a very broad feature as a result of the modification?
The presence of symmetric sp2 carbon as expected in graphite, carbon nanotubes, justify the G-bands as you observed. The D-bands, by another side, appear as structural defects that emerge in the natural carbon system origin, by employing some stress, chemical functionalization, as well as heat treatments, which may generate D-bands if some inclusion is involved.
Those D-bands are forbidden modes, increasing those modes, means less symmetric probed surface.
By your description, the modifications you promote have decreased the degree of the defect in the studied sample, increasing the expected pure sp2 contribution.
Be sure you have not mistaken the width of the D-band with the complete disappearance. That happens often.