In which thread? For significance testing and estimating the overall effect size, the MW and BPB meta-analyses used the unweighted Stouffer Z and unweighted averaging of effect sizes. However, I am following Jessica Utts in using the exact binomial test.
