Nor, for example, my belief that it merges the worst parts of design-by-committee (things that fail to meet an arbitrary and punitive popularity threshold are destroyed forever, never to be given a second chance; things that are popular are not necessarily things that are well-made)
Agreed.
Also, one of the things that strikes me about design-by-committee is that there are better and worse ways of doing it. For instance, I think those elements were generally a boon to the original 5e design (even if I wouldn't have chosen every single outcome myself) because of how they were handled. There was an overall strong vision of what this edition was going to be about from the actual design team, but input was sought from fans of all previous editions. Input was only requested on things that they weren't sure what the best way to do was, and that were in harmony with that vision. Choices were generally made to enable players of as many editions and styles as possible to get what they wanted.
With the 5.5 revisions, I feel like the shift has been more to a popularity contest and trend-following than having a strong vision they are attempting to express in the most well-received way. Here's an example. Any time you ask a modern western audience if they want Option A which is "More freedom, more ability to have it your way, more personal expression!" they will basically
always vote for it, regardless of what Option B is. If you asked people whether they wanted to turn D&D into a system that was entirely build-your-own-class, rather than having defined classes, and presented it using those sort of terms it would win the vote. And of course it would mechanically lose a lot of what makes D&D. (Even though I prefer classless systems, I would not consider such to qualify as D&D at all.)
Here are some examples of 5.0 versus 5.5. Advantage/Disadvantage was a new mechanic they experimented with that didn't really have much relevance to the design vision. It was just a mechanic. It was wildly popular so it got in. Proficiency dice was a similarly tested mechanic. Even though Mike Mearls loves it, it wasn't popular, so he sacrificed his own interest in it and relegated it to a small part of a page in the DMG.
Dragonborn were very popular from 4e, and even though many people from previous editions didn't care for them, there wasn't much reason not to publish them for those who did, so they were in. However, they experimented with some fluff on their origins that was very different from anything that had come before, and it was soundly voted right out and they dropped that bit and made the more traditional version. The fans were helping with the vision. At-will cantrips were similar. Some people don't care for them, but even most of those who were fans of older editions were completely on board with that change, and it felt like it fit the vision of D&D so it made it in.
When in doubt, they tested different things to try to find out what that shared idea of D&D was and ended up with the rules that hit it the best for the most people, while being an obstacle for as few as possible. (When there was no doubt what counted as D&D, like classes and levels, they didn't even ask.)
During the playtest they fiddled with some more abstract mechanics that were popular in 4e (NPCs having little to no overlap with PCs in capabilities), but eventually found people preferred there to be a bit more parity. So in 5.0, monster NPCs used a simplified version of the same spellcasting and such as PCs. They didn't go full 3e style identical construction, but they made them recognizably similar so you could see that such and such is a 5th level cleric, or a 9th level wizard, even if some things were simplified or left out of the stat block. This struck a balance between preferences and stuck with a vision of supporting more play styles.
But in the later revisions, moving into 5.5e, they dropped that and went full on exclusivly 4e-style design, where NPCs are so different you can't easily compare them--which really
only works with the 4e preference. The lead designer (1) says this was to let people "follow their bliss". But unless preferences have completely changed over the intervening 7-10 years (and I have seen little evidence of that), that wasn't what the majority wanted out of 5e. The vision of appealing to multiple playstyles was abandoned in favor of something that some people--including the designer decision makers--happened to really like, even if it seriously impacts the useability for other long-standing D&D play styles. They also started including other 4e-exclusive mechanical abstractions that are playstyle limiting, like having physical weapons change their damage to be entirely an energy type, when the physical weapon is still supposed to be impacting your target in the middle of that energy (so being immune to fire apparently also makes you immune to axes if they happened to be on fire, for instance). They never actually asked about that sort of thing (that I'm aware of). They just pushed it through. It's like some people on the current team
really, really wanted these things that did not make the cut in 2014 for multiple reasons, and decided to push them through now for some reason. That's not how it was done in the 5.0 playtest at all, and my criticisms of these shifts are not an anti-4e thing at all. When I didn't get what I wanted in the 5.0 rules I knew it was because I had a fringe preference that wasn't essential to D&D and might have negatively impacted other people's play experiences. Like, for instance, I was okay with 4e style martial attacks that would do things like push someone without giving them a save--if the attack hit it just happens. But that just wasn't going to work for a lot of people, it didn't really follow the way the game worked for most of it's history, and they were right to go against my preference and implement saves against those things. But the changes they've been making for the last few years don't seem to take any of that sort of thing into account. The design team just has a new direction they want to take things, and unless it is completely shut down by the fans, they keep iterating different ways of doing it until they get something they can label as fitting into acceptable approval, without considering whether it's something they should have even been offering as a part of the core, shared D&D experience in the first place. There appears to be little to no consideration of how the changes might negatively impact some playstyles.
To sum up, the 5.0e playtest interaction between the designers and fans was "How can we make a version of D&D that preserves the enduring elements and feel of the game, and appeals to the most fans new and old possible, including supporting multiple playstyles from past editions?" They had a fixed focus on that and used surveys to figure out how to make it so.
The 5.5e playtest (and changes that were put into prior products without even being playtested) interaction between the designers and fans is "Here's a cool new idea I want to see in my D&D (or an old one that didn't pass muster in 2014)! How can I present and tweak it to get the fans to accept it?" "Oh, and this is like what's trendy, so let's put it in too."
(1) Who I should clarify I actually like as a person and a designer. I just don't think he is the best final decision-maker.