It can be hard to assess the merits of this case but the evidence seems to show that Toyota were pretty sloppy. Having worked on FDA approved software, it's clear that it's reasonable to ship even safety critical products with software defects as long as those defects are independently justified with an explanation of the risks involved. This means that it's hard to condemn a product just because it has defects.

The Toyota case seems to fall well short of this. In my view a damning comment was the independent measurement of stack usage. Toyota's estimate was 41%, the independent estimate was 94%.