This 2020 stem cell review argued against rodent models of human neurodegenerative diseases:

“Neuronal loss is not caused solely by intrinsic degenerative processes but rather via impaired interactions with surrounding glia and other brain cells. Dysfunctional astrocytes do not provide sufficient nutrients and antioxidants to neurons, while dysfunctional microglia cannot efficiently clear pathogens and cell debris from extracellular space, resulting in chronic inflammatory processes in the brain.

Human glia, especially astrocytes, differ significantly in morphology and function from their mouse counterparts. Recent advances in stem cell technology make it possible to reprogram human patients’ somatic cells to induced pluripotent stem cells (iPSC) and differentiate them further into patient‐specific glia and neurons, thus providing a source of human brain cells.

Astrocytes do not efficiently utilize energy resources and cannot provide adequate metabolic support to neurons. A coculture of healthy human neurons with diseased astrocytes impaired neuronal calcium responses to glutamate and γ‐aminobutyric acid (GABA) as compared to coculture with healthy human astrocytes.

Treatment with sulforaphane:

• Normalized basal level glycolysis;
• Decreased basal level Aβ42 secretion; as well as
• Ameliorated inflammatory response to pro‐inflammatory cytokines TNF-α and IL1-β in PSEN1 mutant iPSC astrocytes.

It is essential to make sure that what we see in the dish is the real patient‐specific phenotype. Transplantation of human brain organoids containing microglia into mice could provide a novel tool for drug screening in vivo.”

https://stemcellsjournals.onlinelibrary.wiley.com/doi/full/10.1002/stem.3309 “Metabolic and immune dysfunction of glia in neurodegenerative disorders: Focus on iPSC models”

This review’s thesis seems plausible. However, one problem with in vitro stem cell studies is that they often don’t have a control group.