The definition of an equivalence class

(= quantum state) requires knowledge of the future

This might invalidate the usual objections against
local hidden variable theories ...

?