Improving gRPC benchmark configurations.

[shileiyu]

Is your feature request related to a problem? Please describe.

The benchmark of gRPC framework was collected under less optimal configurations.

1. WRT unary Echo, the client didn't leverage the ability of multiplexing requests over single HTTP/2 connection.
2. WRT streaming Echo, There is only one connection has been established, which may lead to head of line problem very quickly.

Comparing KiteX benchmark with those results may misleading developers.

Describe the solution you'd like

1. At least, explicitly disclose how gRPC benchmarks were collected.
2. It would be better to have a comparison with results collected under more optimal configurations.

[ppzqh]

Thanks for the feedback. I will try to revise the benchmark description and config.

[YangruiEmma]

I just noticed your issue. If you feel that there are problems with the testing code, please feel free to submit PR to help us optimizing it. We would be appreciate it!
Currently, I think the testing methods for Kitex and GRPC are aligned.