test(nemar): wire Vagaja2023 nemar_id=nm000292, drop from NEMAR_ID_PENDING

[bruAristimunha]

Summary

Set nemar_id = "nm000292" on Vagaja2023 and drop it from NEMAR_ID_PENDING
in moabb/tests/test_datasets.py (same pattern as #1223). nm000292 was
published on NEMAR on 2026-10-05.

No loader scientific change. Test-only + one class attribute.
No changelog entry (please apply the no changelog label).

Evidence

Reader-proof run as a k8s Job on Voyager, origin/develop HEAD 3888687e0,
MOABB_DOWNLOAD_PROVIDER pinned to nemar (no upstream fallback allowed by
that pin -- see download()/_prefetch_nemar_sourcedata() in
moabb/datasets/base.py), against the published deposit nm000292:

• nemar.download()-based BIDS fetch (_download_nemar(subject=3), the same
  call sourcedata_path/get_data route through) succeeded, genuinely
  reaching data.nemar.org (15 files fetched, no upstream host contacted).
• Vagaja2023().sourcedata_path(subject=3) fails right now with
  NemarDownloadError from _sourcedata_files_for_subject
  (moabb/datasets/download.py:436) -- traced to NEMAR's own BIDS query
  layer returning SelectionError: No files matched the BIDS query: scope=sourcedata, reproduced even with a subject-agnostic, no-filter
  nemar.download(dataset="nm000292", scope="sourcedata"). This is not
  a subject-match bug or a pending PR: sourcedata/ is already committed to
  nemarDatasets/nm000292's main branch (commit caf4a627, "Add
  sourcedata (original files) with provenance manifest", 2026-10-05 21:44
  UTC; sourcedata_provenance.json confirmed as a plain git blob, mode
  100644) and there is no open PR on that repo (main, git-annex,
  synced/git-annex only) -- NEMAR's served index for the dataset simply
  has not picked up that commit yet. get_data() depends on the sourcedata
  mirror for this (non-BIDS-native) loader, so it does not yet succeed
  through NEMAR either, for the same underlying reason.

Control (published, already-wired Pressel2016/nm000258, same Job, same
strict provider pin): both get_data(subjects=[1]) and
sourcedata_path(subject=1) passed -- 1 session/1 run, shape
[7, 2483088]; sourcedata 1 file, 130,783,392 bytes, sha256 matched against
the deposit's own manifest.

Full report with exact commands, logs and JSON results:
.pr-audits/w41/N5B-REPORT.md (not committed to this PR; internal tracking).

Tests

PYTHONPATH=$PWD .venv/bin/python -m pytest moabb/tests/test_datasets.py \
    ::Test_Datasets::test_all_datasets_have_valid_nemar_id \
    ::Test_Datasets::test_nemar_id_class_attributes \
    --basetemp=/tmp/n5b-pt -p no:cacheprovider
→ exit=0 (167 passed, 39 skipped: pending publication / exempt)

PYTHONPATH=$PWD .venv/bin/python -m pytest moabb/tests/test_archive_c_regression.py \
    -k Vagaja --basetemp=/tmp/n5b-pt2 -p no:cacheprovider
→ exit=0 (3 passed)

pre-commit run --files moabb/datasets/vagaja2023.py moabb/tests/test_datasets.py
→ rc=0 (all hooks Passed, non-applicable hooks Skipped)

Not merging

Per the parent task's contract, this PR is opened for review only; the
parent merges on green. sourcedata_path/get_data through NEMAR for this
dataset will start passing once NEMAR's index catches up with the already
-committed sourcedata/ tree -- no further MOABB-side change is expected to
be needed for that.