Ok. I’ve heard back from the LiteCore team. Neither of us sees anything odd about that log.
It shows a replicator that cannot, initially, find its target. Eventually, though, it does get the connection open. Replication proceeds until all of the documents have been pushed and the replicator goes idle. Sometime later, the local application calls stop() at least twice. The replicator shuts down the connection normally.
It looks to us as if the replication is behaving exactly as it should.