Create how-to optimize transpilation guide

[garrison]

This is a draft. I will update the last section once #517 merges. This PR will probably wait on that to merge first.

Remaining action items

• Review and get ready for merge Add support for generating subexperiments with LO's translated to a native gate set #517 - then update last section here
• Open issue about SingleQPDGates when cutting for depth; link to it in this new how-to - DONE Make cutting-for-depth compatible with SingleQPDGates in the workflow #576
• Try to remove need for SparsePauliOp on following line: SparsePauliOp(subobservables[label]).apply_layout(transpiled_subcircuits[label].layout).paulis
• Request reviews broadly, including from pedrorrivero, once the draft is ready

[coveralls]

Pull Request Test Coverage Report for Build 8941516520

Details

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 95.495%

---

Totals
Change from base Build 8913240803:	0.0%
Covered Lines:	3498
Relevant Lines:	3663

---

💛 - Coveralls

[garrison]

By the way, this how-to guide is essentially an expanded version of the workflow test added in #303.

[garrison]

Try to remove need for SparsePauliOp on following line: SparsePauliOp(subobservables[label]).apply_layout(transpiled_subcircuits[label].layout).paulis

This is tied to the following issues:

• Add apply_layout to other operators Qiskit/qiskit#11824
• Add apply_layout to PauliList Qiskit/qiskit#12302

It is also related to Qiskit/qiskit#12221 but that one is not directly applicable to us because we are not using ObservablesArray (which is something from qiskit.primitives).