Expand perf comparison test and switch to times-based metrics

[Copilot]

The GitHub Actions performance comparison test had limited coverage (6 processes) and used percentage-based metrics that became unhelpful at large speed differences ("99.99% faster").

Changes

• Expanded test coverage: 6 → 15 processes

  • Added: smss, wininit, dwm, RuntimeBroker, SearchIndexer, spoolsv, taskhostw, fontdrvhost, conhost

• Switched to times-based comparison format:

  Before: "C++ is 99.99% faster than Go"
  After:  "C++ is 10x faster than Go"
          "Rust is 2.5x faster than C++"
          "C++ and Go have similar performance"  # when within ±10%
  

• Improved comparison logic:

  • Uses threshold-based equality (±10%) instead of exact match
  • Flips perspective when other tool is faster ("Go is 2x faster than C++" vs "C++ is 2x slower")
  • Added division-by-zero guards

Modified file: .github/workflows/perf-compare.yml

Original prompt

---

This section details on the original issue you should resolve

<issue_title>make performance comparison test longer</issue_title>
<issue_description>make it have more than just 6 tests.
also since the speed difference is so great, showing "99.99% faster" isn't really helpful anymore. so use times. like 10x faster. 12 faster.</issue_description>

<agent_instructions>it's the github actions perf test</agent_instructions>

Comments on the Issue (you are @copilot in this section)

• Fixes make performance comparison test longer #85

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[supervoidcoder]

lgtm