I ignore the "adventuring day" guideline as I think it's meaningless; there is absolutely no consistency in how games are actually played. What matters is how classes compare to each other.
And what matters is how they compare to each other in a variety of situations. Which includes different encounter and rest models. The moon druid breaks what the DMG says the default should be - and is certainly used.
And I strongly disagree that "overcoats of temp hit points" (which is an oversimplified way to describe wild shape) are an inherent problem. To the contrary, I think they offer a different approach to tanking that is tactically interesting. The current 2024 proposal basically just tries to turn the moon druid into a fighter by giving an actual temp HP adjustment and buff to AC. It's boring.
It's more interesting in play than the 2014 Moon Druid and their ability to suffer zero long term consequences by taking all the damage on the overcoat of temp hp.
"You take no long term consequences" is simply the most boring option possible - and the moon druid is the walking embodiment of that. The 2024 proposal might be less distinctive - but it doesn't single handedly destroy tension and break the PHB guidance.
If you say the template druid is non-viable (even for moon) and we know the 2014 druid breaks the default giluidance leading to a horrible experience for some DMs and some playstyles then the 2024 trial is the best option on the table - with eliminating the moon druid entirely coming in second.
And I am going to hard ban the 2014 moob druid as soon as the new PHB is out, joining only the Twilight Cleric and Chronurgy Wizard.
Soaking: using HP like a sponge. This is the current model for low level moon druids: low AC, high HP. Again, passive.
Nope. The 2014 moon druid does not
soak damage. The barbarian soaks damage (I know you call that resistance). The moon druid would have to take damage to soak it. Instead it wanders around with a de-facto force-field
no-selling damage. "Damage" on the animal form simply isn't damage at all because there are literally no consequences. And that is what needs to be changed.
A druid that soaked damage in animal form would be one that converted their daily spell slots to hp, leading to actually taking part in attrition in the way others do, not pulling out 70 new temp hit points every short rest.
I think we are paying too much attention to them being OP for a few low levels; that's not the main problem,
No. The main problem is how they make running the game with the recommended (or indeed any slow) pacing non-viable at the levels that are most likely to harm new DMs. This is why if we can't find an acceptable version they should be removed from the game entirely.
The next biggest problem is the ridiculous overcoat of temp hp. The tough balance curve is a mere footnote in their list of problems.
And I also don't like the current approach, which seems to be to use wild shape to turn them into kinda sorta fighters.
But you won't accept templates and the overcoat of temp hp is truly, gamebreakingly awful at the biggest weakness of 5e - new DMs trying to learn to DM and not getting their game and intended pacing destroyed.
The new version is an attempt at a compromise to keep at least some aspect of what you want. As far as I'm concerned the main problem with the template version is that it was either half-assed or outright sabotaged by someone who wanted to make sure templates didn't happen so produced a crummy version. (See also: brawler fighter for a cool concept so badly implemented it may as well have been sabotage)