Fix undo-label misnomer, tighten drain-nesting/abort-path docs, mark inferred undo behavior as unverified

Renamed the drain's undo point to "Design View FX state" (it restores as often as it parks). Extended the nesting audit to cover a block opening inside the drain's own block. Marked post-fix undo behavior as inference pending DAW checks.

docs/VERIFICATION.md

@@ -48,10 +48,11 @@ Checks for Θ, Ξ, Ψ, Ε, Ρ, Γ, and Ω work that no unit test can close. Buil
 - [ ] Attempt a mode switch while the transport is playing, then while recording — both refuse, visibly (`docs/COMPLETED.md` §"Ψ-W1-T2")
 - [ ] Click the footer mode segment, save, reopen the project — the mode persisted (`docs/COMPLETED.md` §"Ψ-W1-T2")
 - [ ] `[verify — DAW]` Reproduce the strand: project with FX in both modes, switch to Design, save, close, reopen, toggle to Arrange. Toggle back and forth several more times — the arrangement's FX must NOT converge on permanently offline/hidden. A track the tool refuses to park is NAMED in the console, and the same refusal is not reprinted on a later reapply unless the refused set changed (`src/shell/view/view_fx_park.h`'s `decidePark` / `reportRefusedParks`)
-- [ ] `[verify — DAW]` **The undo collapse.** In a project with several FX across several tracks, switch modes once and count the undo points REAPER shows (Ctrl-Z tooltip / undo history): expect exactly TWO — `ReaSampler: Design View FX park` above `ReaSampler: activate <mode> view` — never one per FX (`src/shell/view/view_fx_park.h`'s `fxParkUndoClose`)
-- [ ] `[verify — DAW]` Then press Ctrl-Z once: every FX re-onlines in that one step and the flags stay parked. A second Ctrl-Z rolls the flags and the ext state back. The FX point must sit ABOVE the switch point, never below it and never folded into it (`src/shell/view/CLAUDE.md` §Invariants — the observed undo semantics)
-- [ ] `[verify — DAW]` Open a project saved in Design mode and check the undo history: the load-tick reapply writes no FX state (they are already offline), so it must add NO undo point at all — the discard form. A `ReaSampler: Design View FX park` entry appearing on a plain project open means the discard is not working (`src/shell/view/view_fx_park.h`'s `fxParkUndoClose`)
-- [ ] `[verify — DAW]` Whether `Undo_EndBlock2` tolerates a `ReaProject*` that closed between it and its `Undo_BeginBlock2`. The drain closes its block unconditionally on the project-gone abort path — an unbalanced block would swallow every later action into one point, which is the worse failure — but the SDK header promises nothing here. Reproduce by closing a project tab while a mode switch's FX drain is in flight (`src/shell/view/view_fx_park.cpp`'s `FxParkUndoBlock`)
+- [ ] `[verify — DAW]` **The undo collapse.** In a project with several FX across several tracks, switch modes once and count the undo points REAPER shows (Ctrl-Z tooltip / undo history): expect exactly TWO — `ReaSampler: Design View FX state` above `ReaSampler: activate <mode> view` — never one per FX (`src/shell/view/view_fx_park.h`'s `fxParkUndoClose`)
+- [ ] `[verify — DAW]` Then press Ctrl-Z once: every FX re-onlines in that one step and the flags stay parked. A second Ctrl-Z rolls the flags and the ext state back. The FX point must sit ABOVE the switch point, never below it and never folded into it. Verify the EFFECT, not merely that a point exists in the history — a mask narrower than `UNDO_STATE_FX` covers would still show an entry there while doing nothing to FX state on Ctrl-Z, which would look like a pass and isn't (`src/shell/view/CLAUDE.md` §Invariants — the undo-ordering premise, still `[verify — DAW]` until this item and the next close it)
+- [ ] `[verify — DAW]` Open a project saved in Design mode with every inactive leaf's FX already offline, and check the undo history: the load-tick reapply then writes no FX state, so it must add NO undo point at all — the discard form. This only holds when the saved state was actually fully parked — a first open after tagging a new leaf, an FX added to a parked track while it was offline, or a plugin hand-onlined since the last save all have real FX state to write and legitimately mint one correct `ReaSampler: Design View FX state` point; don't record those as a false failure. An entry appearing on a project that WAS fully parked means the discard is not working (`src/shell/view/view_fx_park.h`'s `fxParkUndoClose`)
+- [ ] `[verify — DAW]` Whether `Undo_EndBlock2` tolerates a `ReaProject*` that closed between it and its `Undo_BeginBlock2`, AND which of two undocumented block-nesting models REAPER implements — a GLOBAL counter (an unbalanced block swallows every later action, in ANY project, into one point) or a PER-PROJECT one (an unbalanced block on a now-dead project costs nothing). The drain closes its block unconditionally on the project-gone abort path on the assumption the global model applies; record which model is actually observed, not just whether the call survives. Reproduce by closing a project tab while a mode switch's FX drain is in flight (`src/shell/view/view_fx_park.cpp`'s `FxParkUndoBlock`)
+- [ ] `[verify — DAW]` **Block-inside-drain.** The nesting audit at `drainDeferredFxParks` only covers the drain opening inside an already-open caller block; the pumping premise it documents (`FxParkQueue::take`) also makes the inverse possible — a re-entrant `applyMode` (an action fired mid-drain) or a re-entrant `OnTimer` -> `bankPanelRefresh` -> `mintManagedLanes` opening ITS OWN `Undo_BeginBlock2`/`Undo_EndBlock2` pair while the drain's block is still open. `g_draining` blocks only a second `drainDeferredFxParks` call, not this. Reproduce by triggering a mode-switch action or a lane-mint mid-drain (a heavy FX chain widens the window) and check the undo history for a split or misordered point (`src/shell/view/view_fx_park.h`'s `drainDeferredFxParks`)
 - [ ] `[verify — DAW]` For a track the console names as left unparked, follow the message's recovery IN FULL: re-enable its TCP/mixer visibility, main send and FX enable, **and set every FX in its chain online** (FX enable is the chain bypass — it does not bring an individually offlined FX back). Then switch modes twice: it parks and restores normally, with no further console line. Do the flag half ONLY on a second such track and confirm it is still refused — the FX half of the trust test, and the reason the message spells the step out (unit-tested only as a decision, never against a live chain)
 
 ## Actions and drops