fix(toolkit-lib): clean up the loser of noOlderThan()'s timer/resolver race

[Adityaj0]

fixes #1869

Reason for this change

BackgroundStackRefresh.noOlderThan() races a "wait for the background refresh" promise against a "reject after ms" setTimeout via Promise.race(), but never cleans up whichever side loses:

return Promise.race([
  new Promise(resolve => this.queuedPromises.push(resolve)),
  new Promise((_, reject) => setTimeout(() => reject(new ToolkitError(...)), ms)),
]);

• Refresh wins (the common case): the losing setTimeout handle is never captured, so it's never cleared -- it keeps the event loop alive and its reject closure retained until it eventually fires on its own.
• Timeout wins: the resolve callback already pushed into queuedPromises for the losing branch is never removed. Only justRefreshedStacks() ever drains that array, so the stale resolver sits there forever -- under cdk gc's polling loop with many concurrent timed-out noOlderThan() calls, queuedPromises grows without bound.

Description of changes

Restructured noOlderThan() around a single Promise executor instead of Promise.race() over two independently-constructed promises:

• onRefresh clears the timeout handle before resolving.
• The timeout callback removes its own onRefresh entry from queuedPromises (via indexOf/splice) before rejecting.

Whichever side wins, the other side's state is always cleaned up.

Testing

Added three tests to test/api/garbage-collection/garbage-collection.test.ts's existing BackgroundStackRefresh describe block:

• noOlderThan() clears its timeout handle once the refresh lands (asserts via jest.getTimerCount(), following the pattern already used elsewhere in this file for the stop() test).
• A single timed-out call leaves no dangling entry in queuedPromises.
• 25 concurrent timed-out calls don't accumulate entries in queuedPromises either.

All existing BackgroundStackRefresh tests (6) and the new ones (3) pass; full garbage-collection.test.ts suite (32 tests) passes.

Checklist

• Unit tests added/updated
• Integration tests added/updated (not applicable — no new AWS resource types or cross-service interactions)
• No manual edits to generated files

---

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache-2.0 license

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 91.10%. Comparing base (7ff50e7) to head (b60cff9).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1870   +/-   ##
=======================================
  Coverage   91.10%   91.10%           
=======================================
  Files          80       80           
  Lines       12205    12205           
  Branches     1742     1742           
=======================================
  Hits        11119    11119           
  Misses       1050     1050           
  Partials       36       36           

Flag	Coverage Δ
suite.unit	91.10% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.