Aggregate data prior to prep_* functions

[AlexAxthelm]

Currently, a lot of code is repeated in each prep function to apply common filters and merges.

pacta.interactive.report/R/prep_trajectory_alignment.R

It'll be more efficient if the data aggregation and filtering were done once before applying plot-specific data wrangling. It would also make it easier to use the prep_* and plot_* functions independently even in cases where no peer or indices data are available. See for example how it is done in prep_data_executive_summary() in pacta.executive.summary.

And the application in workflow.transition.monitor

Migration of https://github.com/RMI-PACTA/pacta.interactive.report/issues/216