Risk calculation in IT software projects [simplified]

LOW:

Our developers know very good the use case and realized similar use cases very often. It exists a well documented standard library in framework for that use case or an inhouse framework.

MEDIUM:

Our developers principally understand the use case and can imagine well to realize it with a standard or inhouse library.

HIGH:

Our developers principally understand the use case, but never realized something similar. They need some none standard not well tested (and even maybe scanty documented) library to realize it Mr. X / Mrs. Y heared in a former company, that some of her co-workers already used an git open source library for something very similar to the most critical path in implementation for our use case. But that git repository looked more like freak made (that's is possible), rather than a well made university project concerning all security and performance issues including soft & hard limits.

(After contacting the author(ess) of git open source project, (s)he means: "I did this 9 years ago, you have to port it to the newest version of the framework and it will work only slacky with over 100.000 main data records! Feel free to implement a caching mechanism and re-deliver it to my git project tree.)

EPIC:

Our developers don't or won't understand that special use case.

They ask: "Are you crazy or on crystal meth?", when receiving the requirements.

There doesn't exist any library in the company's programming language for that. Maybe some international legends like Fefe or Alan Cox realized something similar in a totally different programming language.

IMPOSSIBLE:

Even Fefe, Linus or Alan (Cox) say: "This is science fiction!"