The commercial viability is orthogonal to whether it achieves AGI, which is effectively what "reproducing the human brain" amounts to in this discussion.

And possibly trilions in running costs, not mentioning all the shady training data sources.