In order to analyse requests we need much better logs, with filtering options and (code, time, path etc) and metrics associated with it, e.g. DB request length, processing resources used etc.

This would allow developers to optimise Strapi for more complex API calls and the associated slowdowns that we have been seeing in production.