fix(us): fail-fast timeout on findMyCar location request

[blka]

Problem

HyundaiBlueLinkApiUSA._get_vehicle_location called self.session.get(url, headers=headers) with the session default timeout (connect=10, read=30). If the rcs/rfc/findMyCar endpoint hangs, the coordinator update blocks for up to 30s on read timeout.

The method already has try/except Exception returning None on failure, and the coordinator already falls back to cached vehicleLocation when location is None. So this is purely a fail-faster concern, not a correctness fix.

Fix

Pass a per-call timeout on the findMyCar request only:

response = self.session.get(url, headers=headers, timeout=(5, 10))

• connect=5 (vs session 10), read=10 (vs session 30).
• Only this one call. Other session.get calls in the file keep the session default.
• On timeout, requests.exceptions.Timeout is caught by the existing except Exception → returns None → coordinator uses cached vehicleLocation.
• No retry added — retrying would wake the vehicle and risk 12V battery drain. Fail-faster only.

Tests

New tests/us_findmycar_timeout_test.py (2 tests): asserts session.get is called with timeout=(5, 10); asserts a Timeout exception results in None (graceful). Full suite green. Ruff + pre-commit green.

[cdnninja]

Do we have evidence of this happening?