Until you know the IC type, the prop-delay is unimportant, since you can't gauge anything until you know the pinout. Further, until you know the logic function and pin definitions, you can't know where to apply the inputs and where to find the outputs.

If you don't know the technology, it's because you haven't got the schematic, or because you're working with surplus/reject parts, in which case, unless you haven't quantity to warrant the effort, you should open one of the packages ... yes, sacrifice a component, and look at the die, in which case you'll then know exactly what it is.

If you don't know the pinout, you won't know where to apply your signals.

Discriminating between technologies involves more than just measuring the propagation delays, as (1) they overlap between technologies, and (2) the technologies have different output current ratings and input current requirements.

Discriminating between technologies or simply measuring prop-delay has little to do with the goal the O/P had, which was to determine (a) whether the component in question was functioning as it should, or (b) what the component is. If either is unknown, the prop-delay doesn't matter.

I'm interested in the notion of a decision tree that you mentioned. How can you apply that to a mix of widely varying functions? That would be interesting to see.

RE