Your analysis looks good, I'm in agreement. As I said, I have a bias which is why I disclosed it.
That said, he's nearly our only view into RAI, and I'd love if instead of, as you point out, took the (easy?) road of "how the book reads" he instead took the road of "this is what we actual meant when we designed it" and followed up with Errata to make it so. I think Wizards took the backlash from 4e Errata too far and now err in the other direction.
With the move away from 4e keywords and mechanicism and into "rulings, not rules", I find JC's focus to exactly what is written on the page to be off target and possible off-brand for the spirit of 5e.
Mearls' tweets (sometimes contradicted/corrected by JC) seems to favor RAI and possible RAF, with honesty of "yeah, this is what we meant".
Yeah, I completely share your bias and I'm in agreement with your assessment of JC's focus and Mearls' tweets.
I've said it before, but I really wish that JC would more often answer, "It doesn't matter." Even if he would say something like, "This is how the book reads, but it is highly unlikely to matter in actual play." Like
the whole combination actions thread. In the grand scheme of things, it just doesn't matter. The gap he's closing is
so extremely narrow and nothing being done here can't already be done (especially by characters using two weapon fighting). "But what happens if I do Y and then I can't do X? What does my character do?" Really? Your DM can't figure this out? How about "
nothing"?
I really hate the tendency to believe that rules have to be built to resemble a bad programming language with the force of laws, and I say this as a person whose job is 75% IT professional, 25% state and federal reporting. The tendency encourages people to debug the rules like a programming language, and to pore over the books for "gotcha" loopholes like you're going to put your DM over a barrel and they'll be forced to let you play as
Pun-Pun. It's even worse than just treating them as inflexible legalese because now they get to involve decision trees.
To go off on a tangent, Jonathan Tweet (a lead designer for 3ed) and Rob Heinsoo (lead designer for 4e) made a d20 game called 13th Age. There are numerous sidebars all over explaining why rules were done a certain way, and how to tweak them. So you have direct insight into RAI, and also the "knobs and adjustments" that 5e talked about before it came out. They even include sidebars about where they disagree with each other and how to do things differently. That level of insight into their design process is something I wish we had for 5e, and JC could be our conduit into it - but instead his public face focuses on the words published, not the game intended as designed. I see that as a lost opportunity, even if it is the least confusing for mass consumption.
Yeah, I picked up the core rulebook and the bestiary for 13th Age just because I liked how it was laid out and I liked the ideas it gave me. Same with Dungeon World. I've still never played either -- we tend to play Savage Worlds or some tabletop board game if we're not running D&D -- but I like the systems.
I was so disappointed when the DMG do didn't this kind of thing a whole lot more than it actually did. I want to know why they chose to do a la carte multiclassing instead of hybrid multiclassing. I want sections describing what DMs should consider when they're adjudicating illusions, stealth, etc. I want to know why sometimes poison is poison damage, sometimes the poison condition, and sometimes necrotic damage. I want to know why advantage and resistance are two separate mechanics instead of two degrees of the same mechanic. I want to know what their purpose was with short rests and long rests and the 6-8 encounter adventuring day. I wanted that
lost-to-the-ether Bounded Accuracy article in the DMG as well. Just
tell us your design goals, for crying out loud,
especially in the DMG. It's like they gave us a fully stocked kitchen and told us we could make anything we wanted, but everything in the pantry was packaged prepared food.