The apparent discrepancies between density functional (DFT) and Moller-Plesset (MP2) calculations for the interaction of lithium with graphene recently pointed out by Ferre-Vilaplana (J. Phys. Chem. C 2008, 112, 3998) are discussed. In his calculations, this author used a finite coronene cluster, C24H12, to simulate graphene. We show here that the DFT binding energies of Li to C24H12 are very weak, similar to those calculated by MP2. The weak binding energy is due to the large HOMO-LUMO gap of the coronene, which makes this cluster rather inert. So, the main reason for the earlier discrepancy is that the C24H12 cluster that Ferre-Vilaplana used to represent graphene is too small to properly account for the electronic structure of graphene. In contrast, graphene has zero gap, which allows for a strong interaction with partial ionic character between Li and graphene, as obtained previously in DFT calculations.