Note [Why does Case have a 'Type' field?]

The obvious alternative is
   exprType (Case scrut bndr alts)
     | (_,_,rhs1):_ <- alts
     = exprType rhs1

But caching the type in the Case constructor
  exprType (Case scrut bndr ty alts) = ty
is better for at least three reasons:

* It works when there are no alternatives (see case invariant 1 above)

* It might be faster in deeply-nested situations.

* It might not be quite the same as (exprType rhs) for one
  of the RHSs in alts. Consider a phantom type synonym
       type S a = Int
   and we want to form the case expression
        case x of { K (a::*) -> (e :: S a) }
   Then exprType of the RHS is (S a), but we cannot make that be
   the 'ty' in the Case constructor because 'a' is simply not in
   scope there. Instead we must expand the synonym to Int before
   putting it in the Case constructor.  See GHC.Core.Utils.mkSingleAltCase.

   So we'd have to do synonym expansion in exprType which would
   be inefficient.

* The type stored in the case is checked with lintInTy. This checks
  (among other things) that it does not mention any variables that are
  not in scope. If we did not have the type there, it would be a bit
  harder for Core Lint to reject case blah of Ex x -> x where
      data Ex = forall a. Ex a.