Don't know who is working on NIP-23 for long-form content but the current description on git is a very basic model of what it should become. Two major problems I see with it from a publisher's point of view:

1) the schema/db model: "article" is a container idea for text + stuff + written by one or more people. You might think of "blog post" as text and maybe a photo, which is the basic idea and clearly what someone has gone for so far, but when you start moving up the publishing ladder in time and effort, you need text + multiple media objects + multiple authors/contributors. The media objects might then be displayed as a gallery of photos, or inline photos with descriptions, or maybe someone wants to associate pdfs with their article. You end up with a media_objects table of different upload types (jpg, png, mov, mp4, mp3, wav, pdf, etc, etc) and some way of associating and displaying 1....n objects within a single article AND it needs to be in the display order that the author wants to tell his story, not the way a client wants to do it.

2) clients controlling who sees a publishers content across the network, or which versions get displayed depending on who can be bothered to code what, is going to be a huge red flag for publishers.