The variablity in primary human hepatocytes is indeed high due to inter-individual differences. However, they are seen as "gold standard" because they reflect reality. To overcome the problem of high variabilities, samples of at least 10 individual patients/donors should be tested for an experiment.

The great disadvantage of primary hepatocytes is the fact that they lose hepatic-specific properties (e.g. expression of CYPs) in culture within few hours to days, so there is not much time for performing experiments.

Stem cell derived hepatocytes on the other hand are easier to obtain and to culture, but unfotunately they often resemble fetal, but not adult hepatocytes. Depending on the research question, they might present a suitable model or not. The variability in such hepatocytes depends of the precursor stem cells used - if hepatocytes are always generated from the same iPS line, variability will be low.

In our lab, we established a non-tumorigenic proliferative human hepatic cell line (called HepaFH3), which has some nice features for toxicology testing (e.g. expression of CYPs and possible induction thereof).

If you are looking for Primary Human Hepatocytes, you can check out MilliporeSigma's inventory: https://www.sigmaaldrich.com/technical-documents/articles/biology/lifenet-cryoplateable-cyropreserved-primary-human-hepatocytes.html