docs: fix Lance version and prefetch claims on landing pages

- index.mdx: "Lance v2" → "Lance v1/v2" (both versions supported)
- why-querymode.mdx: "up to 8 concurrent R2 range reads" → "one
  concurrent range read per column" (no fixed cap; count depends on
  selected columns)

Author: teamchong

docs/src/content/docs/index.mdx

@@ -36,7 +36,7 @@ const top5 = await demo
     Zig SIMD engine decodes columns directly from raw bytes. No Arrow conversion, no DataFrame construction.
   </Card>
   <Card title="Multi-format" icon="document">
-    Parquet, Lance v2, Iceberg, CSV, JSON, Arrow. Same API regardless of format.
+    Parquet, Lance v1/v2, Iceberg, CSV, JSON, Arrow. Same API regardless of format.
   </Card>
   <Card title="Edge-native" icon="globe">
     Runs on Cloudflare Workers + Durable Objects + R2. Same code works locally with Node/Bun.

docs/src/content/docs/why-querymode.mdx

@@ -68,7 +68,7 @@ These do real query optimization work:
 - **Predicate pushdown** — filters run inside the WASM engine, not JavaScript
 - **SIMD vectorized decode** — Zig WASM processes columns with SIMD instructions
 - **Coalesced I/O** — adjacent page reads merge into single range requests
-- **Prefetch** — fetch page N+1 while decoding page N, with up to 8 concurrent R2 range reads per page
+- **Prefetch** — fetch page N+1 while decoding page N, one concurrent range read per column
 - **Partial aggregation** — Fragment DOs aggregate locally, Query DO merges
 - **Memory-bounded spill** — sort and join spill to R2 via Grace hash partitioning when they exceed budget