diff --git a/internal/check-selection-kinds.mjs b/internal/check-selection-kinds.mjs
index 380b39240..b1c19d99d 100644
--- a/internal/check-selection-kinds.mjs
+++ b/internal/check-selection-kinds.mjs
@@ -82,7 +82,7 @@ ok(adm1Cases === 0,
   'the evaluator now HAS an adm1 case — if provinces can resolve, stop refusing them and wire them in instead')
 
 // ── 5. THE BUSY STATE SAYS WHICH ACTION IS RUNNING ─────────────────────────────────────────────
-// Driver: "nothing indicates it's doing anything". A disabled button is visually indistinguishable
+// A busy button must say it is busy. A disabled button is visually indistinguishable
 // from an unresponsive one, so the label has to change — and it must be CLEARED in `finally`, or an
 // error leaves the button reading "Working…" forever.
 ok(/busyAction/.test(VUE), 'the busy indicator is gone — a slow packet build will look like a dead button again')
diff --git a/library/components/BackToTopButton.vue b/library/components/BackToTopButton.vue
index 996649281..964d197ae 100644
--- a/library/components/BackToTopButton.vue
+++ b/library/components/BackToTopButton.vue
@@ -3,7 +3,7 @@
  * BackToTopButton — a ↑ that appears once a long list has been scrolled, and sits NEXT TO the
  * layout's floating hamburger circle rather than on top of it.
  *
- * The Driver asked for it "next to the hamburger circle on mobile", for the people-group grid and
+ * It sits next to the hamburger circle on mobile, for the people-group grid and
  * the regions dashboard list. Both scroll their own container, so the container is a prop: this
  * component owns the affordance, never the data.
  *
diff --git a/library/components/poster/PosterWall.vue b/library/components/poster/PosterWall.vue
index 8d6d1589f..161c4aff0 100644
--- a/library/components/poster/PosterWall.vue
+++ b/library/components/poster/PosterWall.vue
@@ -1057,9 +1057,8 @@ useShadowStyles(
         </select>
       </div>
       <div v-show="!isCollapsed(sec.name)" class="grid" :style="gridStyle">
-        <!-- ★ 'has-tape-open' RAISES THE WHOLE CARD while one of its tapes is revealed. Driver
-             2026-09-28 on mobile: the Card deck options tape "is underlapping everything … really
-             bad on mobile". Cause: .pstocks hangs at top:100%, i.e. BELOW its own card and over the
+        <!-- ★ 'has-tape-open' RAISES THE WHOLE CARD while one of its tapes is revealed. On mobile
+             the Card deck options tape slid under the cards below it. Cause: .pstocks hangs at top:100%, i.e. BELOW its own card and over the
              next one, but every .card is position:relative with NO z-index — so they all sit at the
              same stacking level and LATER cards paint over earlier ones in document order. The
              tape's own z-index:2 only orders it WITHIN its card, which is why it looked fine on the
diff --git a/library/maps/whiteboard-map.js b/library/maps/whiteboard-map.js
index be14f5d7e..83cc4e8cf 100644
--- a/library/maps/whiteboard-map.js
+++ b/library/maps/whiteboard-map.js
@@ -1,8 +1,8 @@
 /**
  * whiteboard-map.js — a map engine used as a DRAWING SURFACE rather than as a map.
  *
- * ★ WHY A CHART USES THE MAP ENGINE AT ALL (Driver, 2026-09-27: "snappy, lightweight, lower definition
- * when zoomed out"). A GeoJSON source is tiled per zoom, so zoomed out it draws LESS; label placement
+ * ★ WHY A CHART USES THE MAP ENGINE AT ALL (it has to stay fast, light and lower-definition
+ * when zoomed out). A GeoJSON source is tiled per zoom, so zoomed out it draws LESS; label placement
  * never lets two labels overlap and reveals them in priority order as the reader zooms; hit-testing and
  * pan/zoom come free. An SVG chart keeps every node in the page at every zoom, which is what makes a
  * big one slow. So the engine is asked for a blank surface and the chart draws on it.
diff --git a/library/maps/whiteboard-projection.js b/library/maps/whiteboard-projection.js
index bfba33b7b..91d1aacd4 100644
--- a/library/maps/whiteboard-projection.js
+++ b/library/maps/whiteboard-projection.js
@@ -25,7 +25,7 @@
  * ★ extentFraction DECIDES HOW FAR A READER CAN ZOOM OUT, which is why it is a parameter with a
  * measured default rather than a constant. The engines stop at zoom 0 (the whole square), so the
  * SMALLER the drawing, the more room there is below its opening view. At 0.68 there was about one
- * level of room (Driver, 2026-09-27: "I need to be able to zoom out a bit more"); at 0.17 there are two.
+ * level of room to zoom further out; at 0.17 there are two.
  */
 export function whiteboardScale(extent, extentFraction = 0.17) {
   return extentFraction / (extent || 1)
diff --git a/library/utils/embedBase.js b/library/utils/embedBase.js
index d7512681c..3f1c1ed7a 100644
--- a/library/utils/embedBase.js
+++ b/library/utils/embedBase.js
@@ -184,8 +184,8 @@ export function resolveAssetsBase(profileConfig) {
  * On any host where the tree root cannot be derived from the script url — a dev server, an ESM build,
  * a tree served from a path that does not match `<name>/<name>.js` — base is '' and that expression
  * yields a BARE RELATIVE PATH. Tier 1 then returns it unresolved, and the reader copies
- * `research-map-plus-grid/research-map-plus-grid/index.html` from a page at http://127.0.0.1:5192/
- * (the Driver, 2026-09-28: "the perfect example of how embeddability is not working").
+ * `research-map-plus-grid/research-map-plus-grid/index.html` from a page at http://127.0.0.1:5192/,
+ * a link that works on no other site.
  *
  * ★ AND IT LOOKS RIGHT IN THE ONE PLACE PEOPLE CHECK. A relative share LINK works when clicked, because
  * the browser resolves it against the current page; it is the EMBED that breaks, because an
diff --git a/plugin-designs/my-upg-100-list/upg-100-list.vue b/plugin-designs/my-upg-100-list/upg-100-list.vue
index 41ff02156..bcd7120d3 100644
--- a/plugin-designs/my-upg-100-list/upg-100-list.vue
+++ b/plugin-designs/my-upg-100-list/upg-100-list.vue
@@ -2962,8 +2962,8 @@ onBeforeUnmount(() => {
            So the colour swatches and the hex input are GONE, and the actions that remain BUILD
            the two artefacts this app already knows how to make, from exactly what was picked.
            These reuse printPrayerPacket / printPoster rather than growing a third pipeline. -->
-      <!-- ★ THE BUTTON SAYS WHAT IT IS DOING. Driver 2026-09-28 on mobile: "nothing indicates it's
-           doing anything". batchBusy only DISABLED these, and a disabled button is visually
+      <!-- ★ THE BUTTON SAYS WHAT IT IS DOING. On mobile a tap
+           showed no sign of work. batchBusy only DISABLED these, and a disabled button is visually
            indistinguishable from an unresponsive one — so a reader on a phone taps, sees the label
            unchanged, and taps again. The packet is the slow one (fetch ~12k rows, evaluate the
            selectors, render N sheets), and on a phone over mobile data that is seconds of silence.