It's not because of the molecule you do the k-point sampling, but because of the electrodes. So if you have 1D electrodes (nanotube, graphene nanoribbon, nanowire, etc) then you only need 1x1 k-points, but if it's a metal surface you need to properly sample the periodicity of the surface unit cell.