Can you explain why you need the time of a click? It's a rather strange use case for an automated test.
For a simple, crude implementation, you can record the test client's time before the click() call and the test client's time after the call. You can then average the two times to get a pretty good estimate of when the click actually happened. Your error is determined by the latency of the request from your test client to the test device. If you need even more accuracy, you'll really need to explain why you need the time.
