Say what the day-cycle layer change does to the picture, and where the ratio was measured

Code review r1 found two documentation claims that the task's own contract
contradicts.

Both changelogs promised the picture was unchanged "to the pixel". It is
not: splitting the layer moves the wall hatch by a sub-pixel, and that is
exactly why four day-cycle baselines are re-taken. The entry now says what
a reader can check — the outline around the plan matches pixel for pixel,
the hatch inside the walls is anti-aliased a shade differently.

docs/SUN.md credited Firefox with the fifteen-fold ratio. That number comes
from headless Chromium through CDP tracing; the Firefox profile that opened
the issue measured the same cause differently — 23 MB of texture uploads per
frame and about nine frames per second. Each number now names its engine.

Issue: #532
User-Visible: no

docs/CHANGELOG.md

@@ -6,8 +6,10 @@
   outer outline is a blur over the whole sheet, and it lived in the same layer
   as the plan: any mouse movement made the engine re-run it. The outline now
   keeps its own layer. Measured on a demo-stand pan, rasterization fell from
-  fifteen times a static background to the same order as one. The picture is
-  unchanged, to the pixel
+  fifteen times a static background to the same order as one. The view is the
+  same: the outline around the plan matches pixel for pixel, and the hatch
+  inside the walls is anti-aliased a shade differently — splitting the layer
+  moves it by a sub-pixel
   ([#532](https://github.com/Matysh/houseplan-card/issues/532)).
 
 ## v1.74.0 — 2026-09-11

docs/CHANGELOG.ru.md

@@ -12,8 +12,9 @@
   это размытие по габариту всего листа, и оно жило в общем слое с планом: любое
   движение мыши заставляло движок прогонять его заново. Теперь контур держится
   отдельным слоем. Замер на панораме демо-стенда: растеризация упала с
-  пятнадцатикратной к статичному фону до сопоставимой с ним. Картинка
-  не изменилась ни на пиксель
+  пятнадцатикратной к статичному фону до сопоставимой с ним. Вид прежний:
+  ореол вокруг плана совпадает пиксель в пиксель, а штриховка внутри стен
+  сглаживается чуть иначе — разделение слоёв сдвигает её на субпиксель
   ([#532](https://github.com/Matysh/houseplan-card/issues/532)).
 
 ## v1.74.0 — 2026-09-11

docs/SUN.md

@@ -79,9 +79,13 @@ The public setting remains a two-value selector. `static` uses `bg_color`.
   is kept on its own compositing layer (`will-change: filter`, scoped to the
   day-cycle stage). The group holds paper silhouettes only and does not change
   while the plan is hovered or panned; without the hint every repaint of the
-  plan re-ran three blur passes over the whole sheet, which cost a Firefox
-  window about fifteen times the rasterization of a static background (#532).
-  A static background carries neither the filter nor the hint.
+  plan re-ran three blur passes over the whole sheet. Measured in headless
+  Chromium (CDP tracing, summed `RasterTask`) on a demo-stand pan, that cost
+  about fifteen times the rasterization of a static background; the owner's
+  Firefox profile that opened #532 showed the same cause as 23 MB of texture
+  uploads per frame and about nine frames per second, without a comparable
+  ratio of its own. A static background carries neither the filter nor the
+  hint.
 - Only the environment and the zero-offset alpha-aware outline outside the
   grouped plan-paper footprint change. The plan, paper, floors, room fills,
   Glow/spill, devices, labels, decor/backdrop, vacuum, hover, and window rays