Add case-study vignette for new `model_compare` function

[florence-bockting]

Description

The new model_compare API, as developed in #380, is currently described in the vignette "model_compare.Rmd" (can be found in #380).
The focus of this vignette is

• introducing the new user-facing API
• explaining arguments, and
• providing code examples.

What is missing is an additional vignette that shows the use of model_compare() within a case study. The focus of this vignette should then rather be

• How to use model compare within the Bayesian Workflow?
• What should I do when I see specific warnings?
• How to interpret the output of model_compare