feat: support bin width normalization in data/MC plots

[alexander-held]

This adds a new plot_options dictionary to visualize.data_mc which allows configuring figure options per region. The normalize_binwidth setting takes a reference width (as a number) and a unit (as a string) to then normalize the distribution accordingly. An example looks like this:

plot_options = {"Signal_region": {"normalize_binwidth": (100, "GeV")}}
cabinetry.visualize.data_mc(
    prediction_prefit, data, config=config, plot_options=plot_options
)

to-do:

• ensure that this design is the right direction to go (could also support log settings per region in this fashion)
• tests
• consider also adding x-axis label setting and bin edge values (useful for workspaces built externally), could also be follow-up PR

* add new plot_options setting to allow customizing figures per region in visualize.data_mc
* support bin width normalization via normalize_binwidth setting in plot_options