wpt: add --diff flag that diffs a run against the latest main-branch results

[nicoburns]

Summary

cargo run -rp wpt <suite> --diff now compares the local run against the latest main branch results, so regressions/improvements are visible locally without going through CI.

The baseline is https://dioxuslabs.github.io/blitz/wptreport.json.zst (same artifact the website consumes). It is cached in wpt/cache/ (gitignored) as the raw .zst body plus the response's Last-Modified value, and every --diff run revalidates with If-Modified-Since, so a 304 skips the download. Download failures fall back to the cache; if there's no cache either, the run still completes and only the diff is skipped.

Diffing uses wptreport::aggregate (bumped wptreport 0.0.5 -> 0.0.13, which added aggregate/merge and WptRunInfo::linux_distro). The baseline is filtered to the suites being run, so a partial run (e.g. css/css-flexbox/alignment) doesn't report every other test as removed. Tests whose subtest counts changed are printed with their before/after status, classified as improved or regressed by pass fraction, followed by a summary:

Diff against main (708eb61323e9dda6d57d6f0a667512bb4dfd0068)
=====================
FAIL(3/5) => PASS(5/5) css/css-flexbox/foo.html
ADD  css/css-flexbox/bar.html
---

   0 tests REGRESSED
   1 tests IMPROVED
   1 tests ADDED
   0 tests REMOVED

Link to Devin session: https://dioxus.staging.devinenterprise.com/sessions/ea75dffa87234844b3902d8f14a202c7
Requested by: @nicoburns

WPT results

No changes in test results compared to main.

_{Generated by the WPT workflow.}

[staging-devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR that start with 'DevinAI' or '@devin'.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring