Noting for myself that this whole right truncation section is a bit hard to read and interrupts the flow of the model specification. Feels like it could use a Pyrenew helper function analogous to compute_delay_ascertained_incidence
Originally posted by @dylanhmorris in CDCgov/pyrenew-hew#422 (comment)