Identifying macro conditional independencies and macro total effects in summary causal graphs with latent confounding