perf(bench): record wasm profile tradeoff

[chh-ay]

Summary

• retain raw interleaved benchmark samples for Rust release-profile levels 3, 2, and s, with the shipped wasm-opt -O3 pass held constant
• record artifact hashes, machine conditions, formula/data samples, and two matched render-gate passes per variant
• keep opt-level = 3: level 2 saves 2.08% Brotli but slows resolved 100k formula paths by 5.72–8.46%; level s saves 6.72% but slows them by 8.60–11.71%

Verification

• canonical bun run build:wasm completed and restored SHA-256 d0a73eb34ef002a017900528e0c63c0d5597f13f0d8f58da0c0b91730160eec2
• six matched bench:check runs completed with no statistically detected render regression
• three interleaved full data/formula captures per variant include 500k sort and 1M edit
• no Changeset: measurement artifacts only; no shipped package or configuration change

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
sheetwrite	Ready	Preview, Comment	Jul 27, 2026 1:56pm