- Violates x=x. If you mess with 2000 years of logic, it'll bite back in ways that you don't expect.
- Causes problems for ⍳ ∊ etc. whose definitions are based on =
- Causes problems for ⍋. Where is NaN in the TAO?
- Causes problems for any of the functions with boolean results, because the results of Nan f x and x f NaN are supposed to be NaN. Which is interesting because NaN=NaN is supposed to be false? Whence NaN≠NaN would be true? What about ≥ and ≤? (So for at least 4 functions the results are not NaNs.)