docs: update performance data and clarify benchmark context in README.md

Author: yuriizee

README.md

@@ -8,7 +8,7 @@
 [![PHP](https://img.shields.io/badge/PHP-%5E8.4-777BB4?logo=php&logoColor=white)](https://packagist.org/packages/std-out/simple-data-objects)
 [![License](https://img.shields.io/badge/license-MIT-green.svg)](LICENSE)
 
-**Lightweight, attribute-driven DTOs for PHP 8.4+ — up to 37× faster hydration and serialization than the most popular alternative, zero reflection at runtime.**  
+**Lightweight, attribute-driven DTOs for PHP 8.4+ — up to 60× faster hydration and serialization than the most popular alternative, zero reflection at runtime.**  
 Works standalone or inside Laravel 12–13.
 
 ```bash
@@ -31,18 +31,18 @@ composer require std-out/simple-data-objects
 
 ### Performance
 
-Benchmarked against **the most popular full-featured data-object library in the PHP/Laravel ecosystem** — identical DTO shapes, 20,000 iterations per scenario, PHP 8.4:
+Benchmarked against **the most popular full-featured data-object library in the PHP/Laravel ecosystem** — identical DTO shapes, 20,000 iterations per scenario, PHP 8.4, inside a fully booted Laravel app (not a synthetic standalone script). Medians of 5 runs:
 
 | Scenario | Simple Data Objects | Popular alternative | Advantage |
 |---|---|---|---|
-| Hydration — flat DTO | ~4,500,000 ops/s | ~130,000 ops/s | **~35× faster** |
-| Hydration — nested DTO | ~2,200,000 ops/s | ~74,000 ops/s | **~30× faster** |
-| Hydration — collection of 20 | ~270,000 ops/s | ~7,500 ops/s | **~36× faster** |
-| Serialization — flat DTO | ~7,400,000 ops/s | ~200,000 ops/s | **~37× faster** |
-| Serialization — nested DTO | ~4,000,000 ops/s | ~117,000 ops/s | **~34× faster** |
-| Peak memory — streaming 50,000 rows | 0.26 MB with `lazyCollection()` | ~13 MB | **~50× less memory** |
+| Hydration — flat DTO | ~6,550,000 ops/s | ~132,000 ops/s | **~50× faster** |
+| Hydration — nested DTO | ~3,390,000 ops/s | ~95,000 ops/s | **~36× faster** |
+| Hydration — collection of 20 | ~209,000 ops/s | ~10,200 ops/s | **~21× faster** |
+| Serialization — flat DTO | ~14,900,000 ops/s | ~249,000 ops/s | **~60× faster** |
+| Serialization — nested DTO | ~7,500,000 ops/s | ~166,000 ops/s | **~47× faster** |
+| Streaming — 100k-row CSV import | ~67,200 rows/s | ~35,800 rows/s | **~87% faster**, same flat ~12 KB memory footprint |
 
-Absolute numbers vary with hardware; the ratios stay stable across runs. CPU time per operation follows the same ratios — less CPU burned per request means more headroom per server.
+Absolute numbers vary with hardware; the ratios stay stable across runs. CPU time per operation follows the same ratios — less CPU burned per request means more headroom per server. Streaming a large import with `lazyCollection()` keeps memory flat regardless of row count — the win there is architectural (no full materialization), not a per-row memory difference from the alternative, which also streams comparably once both sides are measured on equal footing.
 
 Don't take the numbers on faith — **[run the benchmarks yourself](https://github.com/std-out/simple-data-objects-benchmark)**: clone the companion repo, `make bench`, or swap in your own payload shapes.
 
@@ -147,7 +147,7 @@ Every worker then starts with opcache-compiled metadata **and** hydration/serial
 
 ```php
 foreach (UserData::lazyCollection($csvRows) as $user) {
-    $importer->process($user); // 50k rows, ~0.26 MB peak instead of ~13 MB
+    $importer->process($user); // ~12 KB peak at 100k rows — flat regardless of row count
 }
 ```