jstests/concurrency/fsm_workload_modifiers/random_manual_migrations.js performs manual moveChunks, and has a list of expected error conditions. However, this list of expected errors is not exhaustive, and chunk migrations may fail with all sorts of error codes while the add/remove shard background hook is running, in the case where a migration recipient is removed (see BF-40386).

We should come up with a way of handling this that does not rely on waiting to see every possible error and explicitly ignoring it.

jstests/concurrency/fsm_workloads/random_moveChunk/random_moveChunk_base.js also has a notion of acceptable moveChunk errors; we should ensure that all tests that derive from these two bases have consistent error handling logic, if possible.