Refactor/Simplify the quantized_op_fusion_pass.py with subgraph_rewriter

[psiddh]

This is a follow up issue to track the feedback on this PR (#13296) and potential future refactor
Simplify this by using a subgraph_rewriter?
see - https://docs.pytorch.org/executorch/stable/compiler-custom-compiler-passes.html#level-2