Change all priors to new dist interface, not just delays

[sbfnk]

          For a future PR this should probably be updated to use `normal()`.

Originally posted by @seabbs in #504 (comment)

This will require moving the conversion to discretised distribution to the processing for stan (which is possibly a good idea anyway).

[sbfnk]

Will be picked up once #504 is merged.

[jamesmbaazam]

The linked comment does not lead anywhere so I'm struggling to get the context. I came here after seeing the code chunk here probably needs to be updated to use the new distribution interface.

Found the comment: https://github.com/epiforecasts/EpiNow2/pull/504/files#r1408162989 (Note to future James: It takes a second to load, so wait a little.)

[jamesmbaazam]

The linked comment does not lead anywhere so I'm struggling to get the context. I came here after seeing the code chunk here probably needs to be updated to use the new distribution interface.

Found the comment: #504 (files) (Note to future James: It takes a second to load, so wait a little.)

I'm paraphrasing the context of the linked comments here because it takes a while to load, giving the impression that the comment is lost:

In the current implementation of rt_opts(), the prior argument is specified as a list instead of Normal(), so @seabbs suggests to update to use the current EpiNow2::Distributions() interface.

@sbfnk says: "Agree in principle though this would probably require extension of the distribution interface to not automatically discretise, or only to discretise when converting to stan code (possibly a good idea anyway?)."