improve legend labels for `plot_causal_impact_sales`

[williambdean]

We can see in the notebook here:

https://www.pymc-marketing.io/en/latest/notebooks/customer_choice/mv_its_saturated.html#scenario-1-own-versus-competitor-sales

that the legend labels are all the same for each color

Hard to distinguish with many like with model3.plot_causal_impact_sales() call