[client] Restore state test

[LayneHaber]

Describe the bug This test only passes because of a special case in the listener error handling, here. The bug is likely a result of listeners not being properly cleaned from the channel provider on restart

When you run the test, you will see the logs:

Reclaiming pending async transfers
[NodeApiClient] Node responded to indra7dvysa8ze5qfqkPLXfZup56omcwtQyccgcW9Xic7TCNZ2fjEEf.transfer request in 17 ms
[ConnextClient] Reclaiming transfer 0x45d96adc2ce5f9c49f19db10c81fd4d6feaadf47d496ea69948fe1079eddd580
[ResolveLinkedTransferController] Resolving link transfer with id 0x45d96adc2ce5f9c49f19db10c81fd4d6feaadf47d496ea69948fe1079eddd580
[CF-ProposeProtocol] Response started
[ConnextListener] Done processing propose install event in 1 ms
[ConnextListener] Done processing propose install event in 2 ms

Notice the two Done processing.. event logs when there should only be one.