[ET][Optimized] Eliminate usage of ET_SWITCH_SCALAR in optimized kernels #4638

Job	Run time
apple (macos) / build	7m 41s
apple (ios-simulator) / build	7m 20s
apple (ios) / build	7m 17s
apple (pybind) / build	7m 0s
linux (llm, linux.2xlarge, executorch-ubuntu-22.04-clang12) / build	6m 14s
linux (llm, linux.arm64.2xlarge, executorch-ubuntu-22.04-gcc11-aarch64) / build	5m 29s
linux (pybind, linux.arm64.2xlarge, executorch-ubuntu-22.04-gcc11-aarch64) / build	5m 51s
apple (llm) / build	6m 25s
linux (pybind, linux.2xlarge, executorch-ubuntu-22.04-clang12) / build	6m 30s
linux (linux, linux.2xlarge, executorch-ubuntu-22.04-clang12) / build	6m 26s
linux (linux, linux.arm64.2xlarge, executorch-ubuntu-22.04-gcc11-aarch64) / build	5m 57s
windows (pybind) / build	4m 22s
	1h 16m 32s

Provide feedback