The Python runner's "test report" is quite simplistic; Bash one too; ditto Rust one

What do we actually want out of a "test report"? What would Subplot-using projects want?

Edited by Lars Wirzenius