Stage 1.10: Fix hard-coded column-header literals (report priorities 1–3)

[jepegit]

Part of Stage 1 — behavior-preserving construction (see the Stage-1 tracking issue). Stage 0: #439. Plan documents live in the shared workspace: cellpy-workspace/architecture-plan/ (the architecture-plan repo; formerly architecture-plan/) (alongside the cellpy and cellpy-core repos). Everything in Stage 1 lands on master with the full suite green; nothing flips user-visible behavior.

Goal

Replace the ~250 true hard-coded column-name literals with canonical headers-class
lookups, in the report's priority order:

1. Journal pages (batch.py:277–292 info_df keys, batch_journals.py "selected",
   helpers.py/collectors.py/batch_plotters.py group/sub_group/label sites) →
   hdr_journal.*.
2. Steps/raw access in ocv_rlx.py (~25 sites) and plotutils.py (~40) →
   hdr_steps.* / hdr_raw.*.
3. Instrument loaders (neware_nda.py rename-dict values, arbin/maccor config
   "power"/"dv_dt" values, post_processors.py groupbys) → headers_normal.*.

Also delete the dead block at easyplot.py:721–739 (references the pre-v8 API).

Why

Native-headers plan Phase 0 prerequisite: after this cleanup, renaming a header touches
header classes only, and the Stage-2 flip cannot be silently broken by a stray literal.
Priority 3 matters doubly — loaders define the raw-table contract, and a header rename
would break them without any test noticing today.

Links

• architecture-plan/hardcoded-column-headers-report.md (the full file:line → replacement tables, §8)
• architecture-plan/cellpy2-native-headers-migration-plan.md (Phase 0.2)
• Independent of other Stage-1 issues; can start immediately.

Acceptance

• Re-running the report's scanner (methodology in the report) over the priority-1–3
  files returns zero unclassified column-literal hits.
• Full suite green (pure refactor — the strings resolve to identical values).