= has a single arity returning true. Should it also have a 0 arity returning true as well?
This would be in line with the results of (and), (or), (every? odd []). And would make (apply = coll) safe in all cases, even when coll was empty.
From slack:
> Agreed - I had an #(->> xs ... (apply =)) in a fdef spec and it was failing in the case when xs is empty, forcing me to write a more verbose function.