I don’t think they skimped on the specials cert. what I think is that they developed the comparability assays there, so that the comparability assays for end product from automation is straight forward. So I inherently completely disagree with your conclusion.
I don't think it's that difficult to prove the end product is the same. Plenty of techniques to do so. Mass spectrometry or HPLC or chemical profile, Flow cytometry for cellular profile, immunoassays for biological profile. Once they get the right programming (incubation, nutrients and waste removal), split samples through manual and automated parallel processing shouldn't take that much time.