docs(tasks): close #66 — 91 orbits, zero silent, and four wrong instruments

The end-to-end gate result, plus the two findings that arrived after the log was
first written: the mixer is clean LIVE (zero "Ardour eats it" across all 10 tracks,
after PLN raised the physical faders mid-session), and the rig does not xrun under
musical load — zero across 60s at maximum event density, scsynth ERR = 1 all session.

Also records the fourth instrument error, which is the best story of the day: the
xrun counter that climbed +816 in 25 seconds and looked exactly like "heavy patterns
cause dropouts on gig week" was probe-chain's own capture streams. Remove the probes,
repeat the identical musical load, get zero. Every probe run creates and destroys ~18
PipeWire nodes and each renegotiation costs Ardour an xrun.

The pattern across all four: the arithmetic was never wrong. Three of them answered a
question about SHAPE with a pair of summary statistics; the fourth measured a system
that included the measurement. What makes them dangerous is that a perturbing
instrument produces PLAUSIBLE numbers pointing at the wrong subsystem, and
confirmation bias supplies the rest — I was looking for a crackle cause and was
handed one.

armada/tasks/018-every-track-plays-every-knob-bites.md

@@ -54,6 +54,26 @@ Two were broken when the day started:
 New tools: `tools/control-lens.py` (the control prover), `tools/check-tracks.sh`
 (the pre-gig gate). Commits `cfa56b6`, `708c2ae`, `7b2e5b4`.
 
+### The gate, run end-to-end at the close of the session
+
+```
+check-tracks: 10 track(s), one at a time
+  bombe_dj: ok — 8/8        wap: ok — 8/8         piment_bresilien: ok — 8/8
+  perfect: ok — 12/12       gimme_acid: ok — 10/10  vague_de_crime: ok — 9/9
+  mafia_sans_serif: ok — 7/7  you_my_sunshine: ok — 9/9  desire: ok — 9/9
+  the_revolution_will_be_sampled: ok — 11/11
+check-tracks: OK — every declared orbit of every track made sound.
+```
+
+**91 orbits, zero silent, zero eaten by a fader.** The mixer count matters as much as
+the track count: earlier in the day, orbits were dying at Ardour's faders on two
+tracks. PLN raised the physical LCXL faders mid-session and the gate now records *no*
+"Ardour eats it" verdict anywhere. (The saved session file still has four faders at
+-inf, so it needs a save — a relaunch would restore the broken mix.)
+
+And the rig is not fragile under load: **zero xruns across 60 s at maximum event
+density**, scsynth ERR = 1 for the whole session.
+
 ## Sel (the shareable learning)
 
 **The instrument was wrong more often than the rig.** Three fixes, and each one is
@@ -73,6 +93,22 @@ cannot actually answer*:
    `mask "<t f!7>"` riser "DECAYING, almost certainly an xfade tail" — twice. Two
    quartile averages cannot answer a monotonicity question. Counting midpoint
    crossings can: one is a fade, two or more is a cycle, and a cycle is a part.
+   Then a *third* false positive: `perfect`'s d4, steady for twelve bins, condemned
+   on the strength of the thirteenth — the capture window had closed inside a gap
+   between events. Fix: ask whether the verdict survives dropping the last bin.
+
+Every one of those three came from answering a question about **shape** with a pair of
+summary statistics, and every fix replaced a statistic with the structural question
+that actually separates the cases. There is a general lesson in there about tooling:
+the bug was never in the arithmetic.
+
+4. **The observer was the effect.** Ardour's xrun counter climbed +816 in 25 s during
+   the stress test — which reads exactly like "heavy patterns cause dropouts", days
+   before a gig, and matched PLN's own report of crackles. Removing the probes and
+   repeating the *same* musical load gave **zero** xruns. Every probe-chain run
+   creates and destroys ~18 PipeWire capture nodes, and each graph renegotiation
+   costs Ardour an xrun. The measurement was the whole finding. An instrument that
+   perturbs the system will happily hand you the conclusion you went looking for.
 
 And the discipline that makes all three survivable: **A-B-A**. On this rig the
 patterns move by themselves — `d2` of `vague_de_crime` plays a different bar every