Fix lifetime scope of the distributed lock coordinator

The fix in commit f6f5fb36915619140a2e32c18aee358549be0b23 addresses some of the issues we were having with multiple instances of the script host generating lease conflices, but the ideal approach would be to change the lifetime scope of the coordinator so we're not creating a new one with every restart and we don't have the pattern where the factory is disposing of the instance it created.

Start by reviewing commit f6f5fb36915619140a2e32c18aee358549be0b23 and tracing the distributed lock coordinator and its factory. The change is complete when the coordinator lifetime avoids creating a new instance on every script-host restart and the factory no longer disposes the instance it created, without the reported lease conflicts.

Written by the indexing model from the issue text.