meta() output

I am conducting a meta-analysis comparing participants with two different diagnoses. All studies in the meta-analysis assess the linguistic construct of interest using both an expressive and a receptive subtask. We are not interested in the expressive vs. receptive contrast, so we would like to run an analysis that includes two (non-independent) effect sizes from each sample and gives us a single summary effect size for the construct of interest, rather than separate summary effect sizes for the receptive and expressive subtasks.

We ran this as a multivariate meta-analysis using meta() from metasem. However, the model output provides separate summary effect sizes for each subtest (i.e., for the expressive and receptive subtasks), which is not what we want. Is there a way to get a single summary effect size (combining expressive and receptive subtasks) for a multivariate meta-analysis using the metasem package? I get a single summary effect size if I use meta3() for a 3-level model, but it is my understanding that a multivariate analysis is more appropriate because the source of non-independence in the data is due to each sample providing more than one effect size. I am also able to get a single summary effect size if I use robu() from robumeta to run a multivariate model, but we have a relatively small number of studies included in this meta-analysis, so RVE is not necessarily appropriate.

I'm not sure if this is a conceptual issue on my part or if there is an argument for meta() that I am missing or what. But I appreciate any suggestions!