Use documented track param parallel_indexing_target_throughput

[wangch079]

The Wikipedia track documented that it supported track param parallel_indexing_target_throughput but it doesn't.

This PR updates the track to use the param parallel_indexing_target_throughput

[gareth-ellis]

I see that this is following the default value in the README, however you should be aware that this will possibly change the results of the benchmark. The number for target throughput is a total requests per second, across all clients, not a percentage. You could add some logic that checks if the parameter has been set, and if so include target-throughput, otherwise exclude it

[wangch079]

The number for target throughput is a total requests per second, across all clients, not a percentage.

This is my understanding as well. If we use the default value for parallel_indexing_search_clients (20), and parallel_indexing_target_throughput (100), each client will try to send 5 requests per second.