Native mid-circuit measurements support composite measurements, etc. (phase 2)

[vincentmr]

Before submitting

Please complete the following checklist when submitting a PR:

• All new features must include a unit test.
  If you've fixed a bug or added code that should be tested, add a test to the
  test directory!

• All new functions and code must be clearly commented and documented.
  If you do make documentation changes, make sure that the docs build and
  render correctly by running make docs.

• Ensure that the test suite passes, by running make test.

• Add a new entry to the doc/releases/changelog-dev.md file, summarizing the
  change, and including a link back to the PR.

• The PennyLane source code conforms to
  PEP8 standards.
  We check all of our code against Pylint.
  To lint modified files, simply pip install pylint, and then
  run pylint pennylane/path/to/file.py.

When all the above are checked, delete everything above the dashed
line and fill in the pull request template.

---

Context:
This PR is a follow-up on #5088 .

Epic 50811 states:

• The deferred measurement principle is not used
• Each shot is a new exploration of the tree of possible mid-circuit measurement results
• The mid-circuit measurement results are correctly exposed to PennyLane's frontend
• Compatible with qubit reuse
• Compatible with qubit reset
• Compatible with qubit postselection
• Compatible with existing support for conditioning in PennyLane
  • Addition, subtraction and multiplication of one or more mid-circuit measurement results

The current PR tackles the requirements above as indicated by the task list.

Description of the Change:

• apply_mid_measure handles post-selection returning a zero state and sample when the state isn't properly normalized (which happens when the state cannot be normalized).
• The concretize method is added to MeasurementValue object to obtain a concrete value from a dictionary of measurement values. This allows handling composite measurements. Other modules like simulate are simplified accordingly.
• Post-selected-zero-state samples are thrown away at the simulate level, handling this exception. This means the final number of samples is lower or equal to the total number of shots.
• Tests for all introduced features.

Benefits:
Mid-circuit measurements are natively supported.

Possible Drawbacks:
Slower in several cases (e.g. when few MCMs).

Missing features:
To do in follow-up PRs:

• parameter-shift support
• optimisation

Related GitHub Issues:
[sc-54987]
[sc-55666]
[sc-55667]
[sc-55668]
[sc-55922]

[mudit2812]

Thanks @vincentmr ! This looks good to me. Very excited to get this in. I'm still a bit worried about the core test runners, but everyone seems okay with it so let's do it

[timmysilv]

one last concern, then I'd say this is good to go!