RFC: Replace the hard-coded list of containers that's used for 'curated content' #6732

One of the data sources we use for Onwards containers is what we call 'curated content'. The content itself is drawn from pressed collections, and we determine which collection to use from a hard-coded list which maps pillars and editions onto container ids.

e.g. an article with the news pillar, coming from a UK edition request, will show 'curated content' from the Headlines container on the UK front.

Generally speaking, these collection ids seem to be fairly stable. But (as far as I'm aware) the collection id is not set explicitly by editors but rather generated by the Fronts tool. So it's inherently pretty fragile to rely on a hard-coded list here, because the collection id could change at any time without anyone realising. This has caused at least one known bug in production, which lasted for many weeks and possibly a lot longer.

Ideally it would be possible for editors to specify which collection should be treated as the source of curated content for a given pillar and edition. However, this would take a non-trivial amount of work on the tools and platform sides, as well as introducing a new workflow for editors. So it would be good to hear opinions on how big an issue people feel this to be.