diff --git a/docs/benchmarks.md b/docs/benchmarks.md
index 1def2ff..161d693 100644
--- a/docs/benchmarks.md
+++ b/docs/benchmarks.md
@@ -23,8 +23,6 @@ shape is visible rather than accidental.
 
 | Bench | Subsystem | Upstream counterpart |
 |---|---|---|
-| `block_encode` | storage | `BenchmarkEncodeColumns` / `BenchmarkEncodeBlock` |
-| `block_decode` | storage | `BenchmarkDecodeColumns` / `BenchmarkDecodeBlock` |
 | `archive_append_live` | ingest→storage | `BenchmarkWriterLiveShape` |
 | `archive_append_backfill` | ingest→storage | `BenchmarkWriterBackfillShape` |
 | `archive_seal` | storage | `BenchmarkSeal` |
@@ -35,9 +33,18 @@ shape is visible rather than accidental.
 | `wire_encode_v1` / `wire_encode_v2` | serve | `BenchmarkDecodeSegmentEvent*` (mirror side) |
 | `cold_replay_batch` | serve | `BenchmarkColdFanout` |
 
-Two upstream benchmarks have no counterpart yet and that is deliberate rather
+The table is exactly what `just bench` prints — ten rows, in that order. If you
+count a different number, the doc has rotted; trust the binary.
+
+Four upstream benchmarks have no counterpart yet, and that is deliberate rather
 than forgotten:
 
+- `BenchmarkEncodeColumns` / `BenchmarkEncodeBlock` and `BenchmarkDecodeColumns`
+  / `BenchmarkDecodeBlock` — bare block encode/decode in isolation. Ours are
+  covered transitively: `archive_append_*` drives encode and `sealed_read_block`
+  drives decode, both through the real call path rather than a synthetic block.
+  Worth splitting out only if a regression lands in encode/decode specifically
+  and the append/read numbers are too coarse to localize it.
 - `BenchmarkDeleteCompactionSyntheticArchive` — compaction is measured, but a
   synthetic-archive harness is a bigger fixture than this file should own.
 - `BenchmarkFlushToTmpfs` — isolates filesystem cost from encode cost; worth
@@ -78,6 +85,12 @@ open costs ~51x an unchecked one. That is why the cold path opens with
 `parseUnchecked` and the admission-critical paths do not, and why upstream
 splits `ReaderOpen` from `ReaderOpenNoVerify`.
 
+("Admission-critical" means a path whose result decides whether a build is
+allowed to ship — see `docs/semantic-parity.md`. Those paths pay for the
+checksum because a silently corrupt segment read there would let a broken
+artifact through. The cold read path is serving traffic that a client can
+retry, so it takes the 51x instead.)
+
 The harness is deliberately naive — a fixed iteration count, no warmup
 discipline beyond a short prime, no statistical treatment. It is here to catch
 order-of-magnitude regressions, not to resolve a few percent. If a number