A value I see in presenting a supplementary model with fewer, commonly available variables across original studies is that it facilitate comparability across studies and maybe (in some cases) improve clarity in meta-analysis by pooling estimates of the same estimand.
