METR has some substantial AI industry ties, so I wonder if those clarifications (especially the one pointing at their own studies describing AI progress) are a way to mitigate concerns that industry would have with the apparent results of this study.