So far, all of the getLastError timeouts in this test have been increased to 2s, except for one lonely getLastError that is stuck at an overly optimistic 50ms and another that's expected to time out. Let's bump that 50ms getLastError to be 2s like all its friends.