Do you have any actual ideas on how to judge quality?
Sure. I'm also aware that they're not at all easy to defend, and may be quite subjective. That's actually part of my point. You are striving to replace the difficult work of analyzing quality with the trivial effort of citing sales figures. Others have mentioned upthread that it is a conversation-ender, not a conversation-starter. "It has to be good, because it sells well" doesn't get us anywhere.
As for your examples, e.g. "design goals," I've already given many criticisms. 5e's rest design is flawed, expecting player behavior that doesn't happen, and Jeremy Crawford has explicitly confirmed this (in softer terms, obviously). It has several flawed classes (Ranger, Sorcerer, Warlock) and subclasses (Beast Master, Berserker) in part because those classes got very little playtesting and essentially zero public playtesting. It specifically advocates three central pillars (combat, exploration, socialization), but fails to actually make every class
directly contribute to these allegedly-central game focuses. Magic has again dominated the power curve of the game, with purely non-spellcaster options a distinct second fiddle, despite explicit designer statements and game mechanics designed to address this gap. It offers if anything a
larger amount of fungible treasure than 3e or 4e, but little to nothing to
do with it.
Advantage/Disadvantage is, exactly as I predicted, rampantly over-used, and as a result a significant portion of gameplay that
could be interesting is instead flattened into "hunt for advantage until you have it," except for those rare and powerful features which stack, such as Elven Accuracy. Feats being (mostly) mutually exclusive with ability score increases forces exactly the kind of bad mechanical decisions people have validly criticized about both 3e and 4e, that dull and bland but powerful bonuses compete with, and are thus favored over, cooler, enriching, flavorful new options. "Bounded Accuracy," beyond being neither
all that bounded nor actually all that much about
accuracy, leaves the game poorly-equipped to create variety of challenge, hence the criticisms, particularly early on, about monsters being largely giant bags of HP with little character.
The books, particularly the 5e DMG, do not IMO do a good job of really actively supporting DMs and players in creating great games, particularly
new players and
new DMs. Instead, they focus on experienced players and setting an extremely traditionalist standard (to the point of almost being restrictive). The explicit deprecation, whether intended or accidental, of various options discourages creativity in a variety of ways. CR is pretty much the same as it was in 3e, mostly useless, you're on your own for figuring out how challenging things will actually be. Skills are a particularly unfortunate subsystem, being a strange hybrid of the limited and relatively closed-ended skills of 3e and the shorter, focused list of 4e, which (IME) brings out the worst aspects of both. Despite the
claimed desire to avoid large growth of power, there's a clear bias toward granting more and more power via spells, and very little more power to anything that
doesn't use spells. And, on the subject of spells, moving to 6 saves, particularly with how poorly saves grow overall, straight-up empowered casters for no reason: it's a lot more likely that you'll have at least
one SoD/SoS spell that can hit a monster's weak saves when there are
six saves to increase and only 2-3 of them will actually "keep up."
This
does not mean it has no positive qualities, nor that it failed to achieve anything at all; both of those would be vast overstatements and completely unfair. It is difficult for me to focus on these because, as I expect the above identifies, I mostly find 5e to be weak in design--not, however, strictly
bad in design. 3e has
bad design, actively opposing its own goals. 5e is by comparison riddled with
weak design, things that
try to achieve an end but fall short, or having two contradictory goals and presenting a tepid realization of both. E.g., early levels being simultaneously option-limited and slow-progressing is meant to make life easier for new players so they aren't overwhelmed, but this runs headlong into the goal of making low levels give the "zero" feel that old-school players value, where life is often nasty, brutish, and short. Trying to fulfill both of these goals leads to an early experience that is
brutally hard and
extremely likely to dishearten or upset new players, and yet is still not sufficiently difficult or simplified for many old-school fans. Neither side ends up well-served.
But, looking at various unequivocally positive design elements, in no particular order:
- Fixing (most) of the "frontloading" problem, classes have a comparatively very smooth curve, such that the à la carte multiclassing system almost always feels like a sacrifice, no matter what level you decide to MC. Given this was a serious flaw of 3e, this is a clear improvement.
- As a corollary: actively making early class levels more approachable (as partially noted above).
- The simplicity of Advantage/Disadvantage in principle (again, in practice I have issues with it, but the idea is unequivocally useful).
- Concentration. An unequivocal improvement in caster balance (relative to 3e, as is the case for many balance-improvement good design choices in 5e), especially since it forces choices rather than being merely punitive or permissive.
- Aesthetics. 4e did very poorly in this regard (other than its splash art), whereas 5e has done a very good job of it.
- Variety. Even if it's not as complete as I would like, they were really trying to be inclusive, and this extends beyond just race and class offerings.
- Subclasses. Note, this is not a mechanic I personally like, but it is unequivocally a design improvement over 3e, enabling faster, lower-overhead manifestation of classic archetypes and concepts (like the "mage-knight.")
- Backgrounds. I myself have advocated for merging 4e's Themes and Backgrounds into what I call Heroic Origins, so 5e Backgrounds are straight-up the same kind of idea translated into a different edition's language. Can't argue with that.
- Finesse as a weapon property. This is a huge improvement (relative to 3e) in both simplicity and usability.
There are probably several others that just don't come to mind right this moment.