Measure change in performance of a single commits

[joemcgill]

Automated performance metrics were added in [55459]. Currently, each time this automated workflow runs, it records two sets of metrics: one for the current change, and one used as a baseline for normalizing results between changes. Currently the baseline measurement is always taken against the 6.1.1 tag in WordPress (the most recent stable release when the workflow was created). In order to better understand the performance impact of a single change, it would be useful to compare directly against the previous commit hash (e.g., trunk) as part of the same workflow, using the same GitHub worker infrastructure to reduce variability between separate runs of the workflow.

Goals of this ticket

1. Collect a set of performance metrics for trunk prior to a commit in addition to the two that are being tested now.
2. Calculate the percentage change between trunk before and after the current change that is being tested
3. Include the percentage change in the workflow summary logs