I disagree that its crummy maths, and I think its designed to create characters with two key strong points and four exploitable weak points.
I don't have a disagreement with the
goal; I do think the math is messed up and things could be better.
The Issues with the DCs become noticeable against very potent opponents, but it makes the ability to engage high level PCs with lower CR challenges much more feasible and interesting.
I think you're mixing two things up here... not sure. I'm trying to keep the latter without the former, although I do think the game starts to break down without using mob rules when you have hordes of lesser foes due to the grind.
Essentially I'm arguing that it's a good thing not to let save DCs get too high, although one might want to control bonuses too. The main reason is to maintain the ability to threaten characters without totally locking others out of being able to do things.
My own experience with D&D in general is that the DCs simply aren't tough enough for most of a PC's career, anyway, so the weak spots a PC has with saves are critical to making any threat feel "threatening" as a result. I suppose an argument could be made that there are other ways to balance this out, sure......but I am not having any issues with the system as it currently stands, and it functions a lot better for my needs than all prior save systems in editions 1 through 4, so it's hard for me to find a point of agreement that the maths are bad in this scenario when they finally, for the first time, feel right.
They're OK as long as you keep DCs to about 18 or under and for much of the game that's how they are. It's only when you start going into the upper reaches of DCs that it starts becoming a problem. I didn't really do the numbers until mid teen levels.
I agree, though, that many DCs are probably a bit too low, while urging people to be careful of very high ones.
Your own example, to use it again, demonstrates that you're trying to work out a contrived method of "fixing" something that isn't broken. The question I raise is: how often are your PCs actually running in to DC 23 saving throws? At what level are these a thing that happens consistently enough to be a major threat? And most importantly, how is it that the PCs have reached such a high level and are still (as a group) unable to resolve this DC? I've yet to see this happen in my games, but I'll concede I haven't run anything higher than level 17 yet.
You start running into those kinds of foes when you get above those levels. You can tell WotC didn't follow their own guidelines written in the DMG because many monsters in the MM have saves that lie quite far from them.
I'm not sure what's particularly contrived about having an effect like dragon breath attack multiple things. There are a number of spells (not enough IMO) that attack multiple saves or do multiple types of damage, e.g.,
Hunger of Hadar,
Ice Storm, or
Flame Strike. This is an elegant way of having effects that get around different kinds of defenses in a partial way.
(EDIT: to be clear, I have seen DC 23s come in to play on rare occasion and they appear to be very nonthreatening to players, especially players with a modicum of cooperation in the group; a PC with the right save will make it, every time, and the ones who don't have that save as a primary will usually fail, sure....but that's clearly the game working as intended, not some sort of accident of design. The fact that the GM can count on this to be a likely outcome is icing on the cake, it makes prepping high level conflicts and having some idea of how they will play out much easier to determine.)
IMO the fact that the
strong save characters make it every time and the others fail every time is
exactly the problem. That predictability for the DM is also predictability for the players. "A dragon, well I'm screwed...." I don't want to have players in that situation. More than once I remember a fight where the poor barbarian's player was reduced to rolling a save for three or four rounds in a row when facing something mind-affecting. That really sucks to be stun-locked for that long. The gap between a strong save character who makes it most of the time and the weak save character who nearly always fails is the issue. At lower levels this isn't nearly so determinative and various things like advantage or disadvantage actually help. When the probability of success or failure is extreme, advantage and disadvantage stop mattering. Have a point of Inspiration against a high DC foe? Why bother? It won't help you. Even many buffs won't help you unless, of course, you built the party to be strong at that. A lot of parties aren't good at buffing. IMO this should not be a requirement but if you want to get to high level play given the structure of the save system, it kind of is.
Challenging the strong save character by raising DCs very high is one reason why DCs have crept up, just as they did in previous versions of the game. It's the same with skills at high levels. Again, don't get me wrong, I want those great wyrms and liches to be tough! Hence having their attacks target multiple defenses. I don't see that being particularly contrived. It's actually playing to the features of the system of being pretty resilient against two attack types and relatively weaker against others and as I said previously there are a number of spells that use this; it's an underutilized approach. A number of spells in the book could be written this way, too, such as
Prismatic Spray.
So my point in the original post is that WotC's math was not as good as it could be, or, to use my stronger word, crummy. WotC often makes math errors to keep things simple but which create a number of potential problems. They did it in 3.X with saves as well. One of the big goals of bounded accuracy was to keep DC creep in check, which is an admirable goal. Unfortunately in the areas of the game where they have binary success/failure, most notably skills (which they didn't spend much time on, or at least chose to leave quite thinly developed, as the case may be) and saves, they didn't really manage. A cure is to keep the DCs (and bonuses) down but have success or failure not be so binary.