fix: do not mask dial errors with KeyError

[enocom]

_remove_cached popped the instance's cache without a default, so a failed
dial against static connection info raised KeyError over the top of the
error that sent it there. The static connection info path builds its
cache per connect and never stores it in self._cache, so the pop always
misses.

Tolerate the miss in both connectors. Only the synchronous Connector
supports static connection info and can hit this today; AsyncConnector
gets the same guard so the two cannot drift.

Related to #449

[enocom]

Here's where we are in this PR Chain:

1. refactor: extract _metadata_exchange helper #566 — refactor: extract _metadata_exchange helper
2. fix: do not mask dial errors with KeyError #567 — fix: do not mask dial errors with KeyError ← you are here
3. fix: close the socket when the metadata exchange fails #568 — feat: classify connection failures
4. chore: add telemetry provider, recorders, and exporter #569 — chore: add telemetry provider, recorders, and exporter
5. chore: add InstrumentedSocket #570 — chore: add InstrumentedSocket
6. chore: record refresh metrics #571 — chore: record refresh metrics
7. chore: wire telemetry into Connector, off by default #572 — chore: wire telemetry into Connector, off by default
8. chore: wire telemetry into AsyncConnector, off by default #573 — chore: wire telemetry into AsyncConnector, off by default
9. feat: enable built-in telemetry by default #574 — feat: enable built-in telemetry by default

Group A (1–3) is independent of telemetry and could ship on its own; #568 carries the only user-visible breaking change. Group B (4–8) is inert because enable_builtin_telemetry defaults to False throughout. #574 flips that default, so reverting it alone turns the whole feature off.