docs: label the performance tables 4.0.0 and fix two stale sentences

docs/introduction/performance.md

@@ -37,13 +37,13 @@ modern-di variant against the rivals whose API matches it, because a single colu
 flatter modern-di against half the set. By-type resolution used to add a fixed lookup cost on
 top of `resolve_provider`: 54-65 ns through 3.2.0, then 21/17/23 ns on C1/C2/C3 once 3.3.0 inlined
 `resolve_provider`'s body into `resolve`. With the template resolver, `Container.resolve` memoizes
-type → resolver directly, and the two columns differ by a few nanoseconds in either direction (242
-vs 239 ns on C1, 146 vs 142 on C2, 760 vs 764 on C3 in the cells below).
+type → resolver directly, and the two columns differ by 3-16 ns in either direction (242 vs
+239 ns on C1, 143 vs 140 on C2, 756 vs 772 on C3 in the cells below).
 
 C4 does not split this way: modern-di's C4 body resolves by reference throughout, while
-dishka and wireup can only be measured by type. That asymmetry cuts against modern-di's
-C4 ratios, though with the by-type surcharge now inside noise, levelling it would not
-move the dishka ratio (1.12) below 1.0. The C1-C3 leveling does not apply to C4.
+dishka and wireup can only be measured by type. The asymmetry is small: the by-type and
+by-reference cells differ by at most 16 ns, under 1% of modern-di's 1.85 µs C4 cell, so levelling
+it would not change which side of 1.0 any C4 ratio sits on. The C1-C3 leveling does not apply to C4.
 
 C6 does not split either, for the same reason: modern-di's C6 body resolves by reference and there
 is no by-type C6 variant to pair against dishka and wireup, so a split would leave that half of the
@@ -59,11 +59,10 @@ for why that matters and which cells it moved.
 
 ## Results
 
-Measured 2026-10-06 with modern-di at the end of the #605 perf series (the 4.0 code, unreleased when measured)
-on an Apple M2 (macOS 26.6.2), CPython 3.14.7 with the GIL, median over 5 runs (ratios paired
-within each run); the footnote under each table bounds the across-run dispersion of each side's
-own median. Rival versions: dishka 1.10.1, dependency-injector 4.49.1, that-depends 4.1.0,
-wireup 2.12.0. Generated by `just bench-report`.
+Measured 2026-10-06 with modern-di 4.0.0 on an Apple M2 (macOS 26.6.2), CPython 3.14.7 with the
+GIL, median over 5 runs (ratios paired within each run); the footnote under each table bounds the
+across-run dispersion of each side's own median. Rival versions: dishka 1.10.1,
+dependency-injector 4.49.1, that-depends 4.1.0, wireup 2.12.0. Generated by `just bench-report`.
 
 The previous publication (3.x, at `630de77`) used the same machine, CPython build and rival
 versions, so the absolute cells are comparable with it as well as the ratios. The
@@ -168,7 +167,7 @@ an async finalizer and closes the child, so it gets the cheaper child build and
 and #605 and pays for #597's lock allocation. The middle column is the previous 4.0 publication,
 before #605.
 
-| | 3.x (`630de77`) | 4.0 before #605 (`f300c2e`) | 4.0 with #605 |
+| | 3.x (`630de77`) | 4.0 before #605 (`f300c2e`) | 4.0.0 |
 |---|---|---|---|
 | C1 / C2 / C3 by reference, modern-di | 253 / 151 / 789 ns | 242 / 146 / 760 ns | 242 / 143 / 756 ns |
 | C4 request lifecycle, modern-di | 2.29 µs | 2.28 µs | 1.85 µs |