Drop the measured speed section

The figure described a build that no longer exists and nothing in the repo
reproduces it.

README.md

@@ -91,18 +91,6 @@ outputs, and CLS norm approximately 1.0. Floating-point rounding can put cosine
 above 1. This single-image check establishes limited conversion parity, not retrieval
 accuracy across datasets. The supplied package retains its original conversion metadata.
 
-## Measured speed
-
-On an M4 Pro running macOS 26.3, native optimized Swift serial inference achieved
-26.6 images/second across 400 images, with median latency 37.39 ms and p95 39.68 ms.
-This was measured on an earlier build that returned only `embedding`; reading
-`patch_embeddings` as well copies a further 2.4 MB per image, which that figure does
-not include. The benchmark used Core ML compute units `.all`, five warmups, and three
-passes; throughput uses the median pass duration. Input feature creation, prediction,
-and vector extraction were timed. Photo fetching, resizing, model loading, and
-database writes were excluded. These are local measurements, not a hardware-independent
-guarantee. Only aggregate timing results are included here.
-
 ## Attribution and license
 
 Base model and DINOv3 research by Meta: