E4: delete batch_tools/ + collectors elevated-kwarg shims

[jepegit]

Epic E — Shim removals (BREAKING) · Delete batch_tools/ + collectors elevated kwargs.

Acceptance

• Delete cellpy/utils/batch_tools/ (incl. dead stubs ImpedanceExperiment, stub engines, batch_reporters.py, OriginLabExporter, do2, _init_old/_new).
• Delete old collectors' elevated-kwarg shims and plotter functions/fig-IO/templates (B4 handover done).
• Keep cellpy.utils.batch / cellpy.utils.collectors as permanent one-line re-exports of cellpy.batch / cellpy.collect.

Depends on A7 + B3 (replacements must exist first). Verifies #668 is resolved. Part of #696.

[jepegit]

Blocked — needs a native DB-journal port first.

Scoping E4 during the Epic E push: batch_tools/ is not dead. Its collectors elevated-kwarg shims are already gone (removed in B4, clean break), but the DB-journal machinery is still the live implementation:

• cellpy.batch.db.journal_from_db (which backs Batch.from_db) delegates the entire database parse to LabJournal.from_db() in batch_tools/batch_journals.py (1134 lines), which drives the Excel/JSON db readers via batch_tools/engines.py (412 lines). db.py's own docstring says it: "LabJournal still owns the actual db parsing (deleted at E4)."
• Tests still exercise LabJournal, batch_helpers, engines, sqlite_from_excel_db.

So "delete batch_tools/" requires porting ~1500 lines of DB-journal parsing (LabJournal.from_db + engines) onto the native cellpy.batch stack first — a real feature port, comparable to C2, not a mechanical deletion. The "A7+B3 done → unblocked" premise covered the facade + collectors, but the DB-journal path was never natively reimplemented.

Recommendation: treat the native DB-journal port as the actual content of E4 (or a dedicated arc), and decide whether it lands in 2.1 or moves to 2.2 (like C2). The cellpy.utils.batch / utils.collectors re-export shims stay regardless.