Dismissing the papers as lies is very different from considering your usage of the studied tools/materials, and deciding you're not actually recreating the failure-modes studied in the papers enough to worry about your current tools.