Migrate elementwise_util callers to the variants with out_dtypes in template arguments (#9841)

This is necessary to take advantage of the previous PR #9388, which
creates dtype-specialized implementations for the non-mixed dtype case.

The size cost of this approach looking at `__TEXT` of
size_test_all_optimized_ops built with bash
test/build_optimized_size_test.sh is 104212 bytes, a 2.2% increase.