@TMeadows (WMF contractor), I was just reading the May 1 status update. I understand that the ideas being discussed are conjecture and spit-balling. Nonetheless I'd like to ask some questions, even if it only serves to help the team focus on key questions to better define and categorize the various design options.
You wrote: it seems the best approach might be to have a system whereby wikitext - for those who want to use it - will be an "under the hood" feature, accessible for more advanced users, with a "surface" interface more user-friendly for lay users or simply those that have no desire to use wikitext.
I really want to say that sounds promising. However I would be unsurprised if those words were literally copy-pasted from some original Flow announcement. One big issue with Flow was that the project manager's idea of "wikitext" was not compatible with what the community believes that word means. When the Flow project was first announced, many editors gave dire warnings that Flow was going to fail unless it had genuine wikitext support.
My first question: Have team-discussions considered that the definition of "wikitext" is an important design question?
My next question is whether the team has an idea in mind on that design point. When the team suggests "wikitext", does that mean that discussions would continue to use the current PHP Parser? (Providing genuine and accurate rendering of article-content in discussions.) Or were they thinking of switching discussions to Parsoid, like Flow?
My next related question: Were they suggesting that new features would be added on top of existing wikitext pages? Or did they have in mind the Flow approach, trying to replace existing talk pages? I think everyone would agree that trying to replace existing talk pages is the higher-risk approach. People will call that Flow2.0, putting the project in a deep hole on day one.