Add parameter summaries to the output

[markjrieke]

would be useful to add poll model parameter outputs to get a sense of where the model is detecting bias:

• beta_b: state-level polling bias (only report for Harris)
• beta_g: group (population) bias
• beta_m: mode bias
• beta_c: candidate-sponsored (partisan poll) bias
• beta_p: pollster bias

join to the mapping tables so you get natural names out (i.e., bias of Marist rather than bias of pollster 3 or whatever)

[markjrieke]

Closed w/2.9