All saving throw attacks scale at a rate of "They don't scale past grade V at all." With how easily PCs can get access to feats, prizes, and raw level bonuses to saving throws it can make any type of combat that isn't a boring grindfest of HP hard to balance at best and completely broken at worst.
Grade I should be something only people with a weak stat or debuffs should worry about at any level, Grade V should be something you want to avoid at all costs. That's how it works with all the other scaling with I being pathetic and X being completely unfair unless you bypass that stat entirely.
Isn't that more or less how it already works, though? Grade I can potentially be threatening for low-level characters with a weak bonus to that save, but even for them it's a minimum 50% chance of success unless you actually have a negative modifier to the roll. Grade V is tough to beat even at level 20 -- you can do it if you really work at it, stacking up a strong base save bonus with every other bonus you can get your hands on, but the same is true of (for instance) trying to hit an enemy with Defense X.
In fact, if you took the quick-and-dirty intuitive route to scaling save DCs, making them work on a I-X grading like other scaling traits and calculating the DC as 10 + the bonus listed for the grade on the "Resilience, Competence, and Health" table, you can see how well the non
-scaling DCs fit with the way save bonuses progress. Setting our benchmarks for comparison at I, III, V, VII, and X corresponding to grades I-V respectively, to account for the difference in granularity, you actually get the same DC at level 20 as the high-grade fixed DCs, and the midpoint grade is pretty close as well (DC 22 for scaling grade V vs DC 20 for fixed grade III). Scaling according to save bonuses would actually make the low grades scarier, with the DC being ~5 higher at level 20 than the fixed grade. At lower levels, the mid grades are closer to their static counterparts, while the high grades are lower. In other words, relative to save bonuses as summarized in the NPC traits table, the fixed DCs sit pretty neatly at values that are fairly easy for the low grade attacks, tough for the high grades, and in the middle for the middle.
Also worth noting is that the DC 30 for a high-grade save attack is pretty comparable to what a PC could manage for the save DCs of their effects by level 20 if they built for save DC (within reasonable limits). For instance, a Mage who loads up on Spellcasting feats in all his feat slots (including the bonus feat from his Specialty) and focuses strongly on Cha (taking a 16 to start out after all origin modifiers, and adding 8 between the usual increase at every 4th level and the three boosts from Arcane Might) would have a spell save DC of 30. Which matches up with the DC of a grade V save attack, and with the average result an NPC with Resilience X will roll (before modifiers from attributes or the like).
Basically, it seems to me that the fixed save DCs are pegged at values that essentially fit the same purpose of "relevant at all levels" that scaling trait bonuses do, without actually scaling. It is in fact precisely as Morgenstern describes it upthread: Save bonuses scale slowly enough that you don't need
DCs to scale to keep pace. I wouldn't say there's any sort of design failure to the DCs being static, just that it's a design decision that aesthetically might seem out of place alongside all the other scaling things. Saves work on fairly restrained scaling paradigm a la 5e's "bounded accuracy", while other numerical values hew closer to 3.X's continuous escalation. Given that both save bonuses and save DCs work on the same paradigm, it's not really a game balance issue that this paradigm is different from how other things work.
(As an aside, I would say there's a bit of a design flaw in save attack DCs and duration/magnitude being locked to each other, but that has nothing to do with whether save DCs are static or scaling.)