While the conclusions make sense, it's weird that we have gotten to a point of meta-analysis of existing "systematic and meta analyses". How much further remove

While the conclusions make sense, it's weird that we have gotten to a point of meta-analysis of existing "systematic and meta analyses". How much further removed from the data can we go without the "error" accumulating and making any such meta analysis useless?