Complete cell examples, synapse communication, and network runtime APIs

[TLDSZ]

Description

This PR integrates a broad set of multi-compartment modeling updates across cells, synapses, runtime scheduling, and network APIs.

The main goal is to make BrainCell more complete for end-to-end workflows: build biologically detailed cells, validate them against NEURON
references, connect them through synapses, and run small recurrent/network simulations through a first-class braincell.network interface.

Cell and mechanism updates

• Completes and refreshes cerebellar cell comparison workflows for PC, GoC, GrC, DCN, BC, SC, and IO models.
• Adds or updates BrainCell/NEURON paired implementations, parameter files, debug workflows, and runnable notebooks.
• Improves multi-compartment runtime behavior for point mechanisms, channel current ownership, clamp routing, and compartment-level scheduling.
• Adds support for additional channel/ion variants used by the comparison examples, including nonspecific ion handling and updated cerebellar
  channel implementations.
• Adds generated morphology and native parser resources for DCN workflows.

Synapse and communication updates

• Adds synapse comparison examples and runners for validating BrainCell synapse behavior against NEURON.
• Improves runtime communication between cells, point mechanisms, synapses, and state buffers.
• Extends event/input routing so synaptic inputs and point clamps can participate consistently in the multi-compartment runtime.
• Updates comparison utilities and fixtures around synapse and channel workflows.

Network runtime and APIs

• Adds a new braincell.network package with first-class APIs for:
  • populations
  • explicit connections
  • projections
  • topology helpers
  • connection lowering
  • delayed event delivery
  • network execution
• Supports fixed-step delayed synaptic event delivery with per-edge/per-contact weights and delays.
• Adds backend selection for event delivery, including scatter fallback and optional brainevent integration.
• Adds network runtime/topology tests and design documentation for the new network layer.
• Updates multi-compartment network examples and benchmarks to demonstrate the new APIs.

Docs, examples, and organization

• Updates docs and notebooks for cells, mechanisms, filters, visualization, and network workflows.
• Moves and reorganizes NEURON comparison examples under examples/neuron_compare.
• Adds design notes for the network implementation and refreshes TODO/changelog entries.
• Updates converter and validation utilities used by the example/model comparison workflows.

How Has This Been Tested

Tested locally with targeted and module-level regression runs.

Targeted regression for the recent network delay and implicit solver fixes:

python3 -m unittest \
  braincell.network.runtime_test.NetworkRuntimeTest.test_per_edge_heterogeneous_delays_arrive_at_different_steps \
  braincell.quad._implicit_test.ImplicitEulerLinearTest.test_one_step_lies_in_cn_bracket

Network runtime and implicit solver module tests:

/home/swl/anaconda3/envs/braincell/bin/python -m pytest \
  braincell/network/runtime_test.py \
  braincell/quad/_implicit_test.py \
  -q

Latest local result for the module-level regression run:

45 passed

Additional coverage in this branch includes tests for:

- multi-compartment runtime state buffers and scheduling
- channel and ion current ownership behavior
- cell construction and runtime initialization
- network topology helpers
- connection lowering and delayed event delivery
- synapse/point mechanism behavior
- NEURON comparison helper workflows and fixtures

## Types of changes

- New feature (non-breaking change which adds functionality)
- Bug fix (non-breaking change which fixes an issue)
- Documentation (non-breaking change which updates documentation)
- Refactoring (no functional changes, no api changes)

## Checklist

- [x] Code follows the code style of this project.
- [x] Changes follow the CONTRIBUTING guidelines.
- [x] Update necessary documentation accordingly.
- [x] Lint and tests pass locally with the changes.
- [ ] Check issues and pull requests first. You don't want to duplicate effort.

## Other information

This is a broad integration PR. Suggested review order:

1. Review the new braincell.network API and runtime behavior first.
2. Review multi-compartment runtime changes around point mechanisms, synapse input routing, and state buffers.
3. Review cell comparison examples and NEURON paired implementations.
4. Review docs/notebooks and generated comparison artifacts last.

The highest-risk areas are:

- delayed event delivery semantics and delay quantization
- runtime state-buffer ownership across cells, synapses, channels, and ions
- compatibility of comparison notebooks and NEURON reference workflows
- backend-specific event delivery behavior when brainevent is available

[sourcery-ai]

Sorry, we are unable to review this pull request

The GitHub API does not allow us to fetch diffs exceeding 300 files, and this pull request has 407