A more complete web benchmark might be one which connects to the database, does some queries, and renders the result. Then we'd look at what percentage of the t

A more complete web benchmark might be one which connects to the database, does some queries, and renders the result. Then we'd look at what percentage of the time is spent in the runtime of each language, and apply Amdahl's law to see if switching is cost effective.