Revert "fix(blob): do not remove existing reports from output dir"

[Yury Semikhatsky (yury-s)]

Reverts #43056

[github-actions]

Test results for "tests 1"

2 flaky

⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:257 › third party 'Partitioned;' cookies `@firefox-ubuntu-22.04-node20`
⚠️ [firefox-library] › library/browsercontext-cookies-third-party.spec.ts:470 › top level 'Partitioned;' cookie and same origin iframe `@firefox-ubuntu-22.04-node20`

52642 passed, 1270 skipped

---

Merge workflow run.

[github-actions]

Test results for "MCP"

1 failed
❌ [firefox] › mcp/cli-core.spec.ts:57 › click link @mcp-windows-latest-firefox

9034 passed, 1495 skipped

---

Merge workflow run.

[github-actions]

🟢 Playwright bot verdict: the failure is a pre-existing flake — this PR is clear

This PR only reverts blob-reporter output-directory handling. The failing MCP test does not exercise that code, and its Windows/Firefox timeout is already documented across unrelated runs.

Details

Overall: the only real failure is the MCP click link test on Windows Firefox. The two library results are retry-rescued flakes, so they do not affect this verdict. Nothing in the three-file blob-reporter diff reaches MCP click dispatch or Firefox navigation.

Pre-existing flake / infra

• [firefox] › mcp/cli-core.spec.ts:57 › click link (mcp-windows-latest-firefox) — the click timed out after 5000 ms while Firefox was performing the action on <a href="/hello-world">, exactly matching the known flake documented in fix(test): unflake MCP cli-core 'click link' on Firefox #42306. Its aggregated history there shows 24 failures in 584 runs (~4%), all with the same click-action timeout, including unrelated runs such as 30014511758 and 30046492910. This PR changes only packages/playwright/src/reporters/blob.ts, its tests, and reporter documentation, so it cannot cause this MCP/Firefox timing race.

_{Triaged by the Playwright bot - agent run}