Also keep in mind that the quantification of height by AFM at these scales is affected tip-sample interactions. Typically, you make a cross-section between your 2D layer and the substrate. If the substrate is SiO2/Si then you already have a difference in the tip-sample interaction between the SiO2 and the 2D material. This difference in sample interaction due hydrophobicity, electrostatics, or van der Waals could mess up the quantification of the height difference you obtain.