Update on "[Executorch] Refactor op_mul's broadcasting utils"

Summary:
Refactoring broadcast handling utils that were added for op_mul. This is
in prepartion use these utils to handle broadcast for other ops such as
add, sub, div.
Plus remove a redundant test

Test Plan:
optimized_kernels_test in CI

Reviewers:

Subscribers:

Tasks:

Tags:

cc larryliu0820 manuelcandales

Differential Revision: [D69491816](https://our.internmc.facebook.com/intern/diff/D69491816)

[ghstack-poisoned]

Author: kimishpatel

kernels/optimized/cpu/op_mul.cpp

@@ -130,7 +130,6 @@ Tensor& opt_mul_out(
           out.numel());
     });
   } else if (selected_optimized_path != ElementwiseOptimizedPath::kNone) {
-    ScalarType out_type = out.scalar_type();
     ET_SWITCH_REALB_TYPES(out_type, ctx, "mul.out", CTYPE, [&]() {
       auto mul_lambda = [](auto x, auto y) { return x * y; };
       return torch::executor::handle_broadcast_elementwise<CTYPE>(