The manual timing calculations were adding Node.js processing overhead
and were less accurate than the native timing provided by the respective
libraries (ping, got, etc.). Now using response.time from each library
for more accurate network response time measurements.