Where's the line for material non-public information, vs something not obvious to most of the public, but discoverable through research? Any engineer sufficien

Where's the line for material non-public information, vs something not obvious to most of the public, but discoverable through research? Any engineer sufficiently motivated to do so could have examined a CPU at any time in the last ten years, and, through skill, access to publicly available information (the instruction set, cpu datasheets) and immense amounts of hard work, uncovered the Meltdown flaw, without any access to Intel's internal trade secrets.