Agreed. Skimming Humbird's analysis, he mentions concentrations of catabolites as a significant limit on cell density, and points out that their removal is usually the job of the kidneys. To me that immediately raised the question of how to design an artificial kidney-like structure that can also live in solution. Similarly, the cleanroom conditions are very difficult to sustain, but what if we could engineer a replacement for the immune system to police the reactors?

Both of those are of course complete science fiction currently, but they're not "thermodynamically impossible" like he seems to suggest. They're 'just' conditioned on a significantly deeper understanding of biochemistry and genetic engineering than we currently have.

Given the current state of the technology and the implications of meat for global warming, I suspect that meat might just become more expensive until it stops being eaten entirely. And when the technology exists to produce it artificially, there won't be a market for it anymore. Speaking as someone who eats meat regularly, it's mostly a matter of conditioning. I don't think I would have independently invented the idea of killing and consuming an animal if other's hadn't taught it to me.