#4416 run doctor --fix --server cannot settle damaged runs left by a crashed serve after restart

swamp run doctor --fix --server cannot settle a run left behind by a serve process that crashed, once serve has restarted.

The doctor handler decides whether a run's owner is gone from the tracker row and the server's instance id. A restarted serve has a new instance id, and the check does not treat a row from an earlier instance as settleable in this path. So a damaged run from the crashed instance stays reported and unsettled over --server. The local CLI, swamp run doctor --fix, run in the repo, does settle it.

Found while fixing swamp-club#2517; that fix shipped with this limit stated.

Expected: after a serve restart, run doctor --fix --server settles damaged runs whose owner was an earlier serve instance that is gone, or tells the user to run the local command.