I'd argue Turing completeness is bad also, Rices theorem and all. But like with macros, sometimes you just need to accept and deal with the badness because the

I'd argue Turing completeness is bad also, Rices theorem and all. But like with macros, sometimes you just need to accept and deal with the badness because the alternatives are worse (or undiscovered).