Enable plotting of model prior predictions

[drbenvincent]

At the moment we have no in-built plotting capabilities to plot the model prior predictions. Adding this would be very useful.

The most obvious thing we could do is to use the existing plot methods, but simply plot from the prior, not the posterior. This might require some changes to the experiment classes so that we have a prior version of everything to plot. This would be an interesting approach because it would allow you to see the implied estimated treatment effect based on the priors you've specified.

NOTE: This may well be easier to implement after https://github.com/pymc-labs/CausalPy/pull/524 because we'd have more consolidated idata objects.