Performance measurements

[ntoll]

In our recent community call (https://www.youtube.com/watch?v=YBkJGn7qJOg) we discussed ways in which we measure and record changes in the performance of different aspects of PyScript. This issue was created as a place in which we can discuss and come to a consensus.

I'd be interested to know:

• What are we measuring? (Speed, resources with different interpreters / network performance characteristics?)
• How we're measuring? (What tooling allows us to gather such metrics?)
• Why we're measuring? (Ensure minimal regressions, what change boundaries cause warnings etc..?)
• When we're measuring? (Manual kick off via make perftest, as part of of CI process etc..?)