While we consider the way forward for case sensitivity we have chosen to remove the existing support for normalizers on the wildcard field.
They may be supported again in future but we'd like to keep options open if we decide on a different approach.