I'd agree with both of these things but faulty data...that's prevalent everywhere. Research, government studies, manufacturing, that's not necessarily a shortcoming of Lean so much as a shortcoming of a person's ability to ensure data quality, which is arguably the harder part of the equation. As for human behavior, that's a tricky one. Plus, it's near-impossible to ensure consistency across people doing the same exact thing. Which is frustrating, however, I do think it can be successfully implemented if data quality and the inclusion of operators and such into it's implementation are accounted for. It shouldn't just be a tool of the engineer or equivalent.Lean suffers badly from bad data and bad incentives. It fails to account for people naturally wanting to cheat and cut corners thereby requiring the non value added steps it so boldly try's to eliminate.