As part of the measurements to measure the impact of MinT for wiki Readers (T373862), we may want to consider the translation response time.
Since slow response times can add noise to the perceived usefulness of the feature, we may want to instrument the response times. In this way, we can identify if results are different for those users experiencing slower translation times.

As part of this ticket, we need to instrument the response times and define the best way to analyze them (e.g., cases where the response is over 3 seconds...).