Based on some ongoing conversations around making use of class DC, I think a few elements jump out:
There are vanishingly few mechanics that exclusively use class DC, and when they do, it's usually for things that can be done with spells, such as crit spec applying a condition with a save.
Class DC and spell DC are often used interchangeably, most notably for many ancestry and archetype feats that use a DC. This ends up creating a lot of extra wording each time.
The very few classes with a legendary class DC tend to be quite well-suited to doing things casters can do: for instance, the Kineticist has access to a lot of AoE and utility, and can even use their class DC to cast many spells via magic items.
In the past, I wondered what would happen if class DC and spell DC were merged, and tried out this sort of change at my table, especially when playtesting content for Starfinder. Specifically, here were be the changes:
Spell DC is no longer a statistic. Instead, any save uses your class DC.
You get just one class DC, and can use this for any DC that requires a class DC, including class features from a multiclass archetype that specify another class's DC.
Classes with a spellcasting feature start out trained in class DC (this is standard on remastered classes, but not classes that haven't yet been remastered, like the Magus), and get proficiency increases to their class DC when they would normally get an increase to their spell DC.
Optionally, classes with a spellcasting class feature also have their spell attack proficiency decoupled from their class DC proficiency in a manner outlined below:
Decoupled Spell Attacks:
To decouple spell attacks from class DC, apply the following, based on Mark Seifter's proposals for spell attack accuracy:
1. All classes with a spellcasting feature become experts in spell attacks at level 5, and masters at level 13. Nobody becomes legendary.
2. Spell attacks benefit from item bonuses to attack rolls, as well as attack potency bonuses when using Automatic Bonus Progression. Use the item bonus from the strongest weapon potency rune on an item you're wearing or holding.
3. Every named staff gets a weapon potency rune appropriate for its level.
4. If the staff would be cheaper than its weapon potency rune, increase its cost to match its rune. This generally would require only minor adjustments (level 10 staves would need their cost bumped up by 15 to 35 gp, some level 16 staves would need their cost bumped up by 35 to 435 gp). If the staff's Price already matches or exceeds that of the rune, no change.
5. Prevent the weapon potency rune from being transferred out of the staff for reselling: if the staff loses its rune, it loses the ability to be prepared and cast spells until it regains that rune once more.
6. If a player wants to craft a personal staff, have the resulting staff come with a weapon potency rune appropriate for its level. If the player has one such rune and wants to supply it during the crafting process, deduct its cost from the total crafting cost as normal.
7. You can apply the same change to Kineticists, having their impulse attack proficiency increase to expert at level 5 and master at level 13 (and no longer legendary at any level). You can even change their impulse attack proficiency to spell attack proficiency, and have their impulses work as spells (so you'd be Casting a Spell). If you do all of this, increase the item bonus provided by a major gate attenuator to +3.
8. As an optional additional rule, you can state as a baseline that spell attacks always use your key attribute as their attribute modifier, with classes and archetypes no longer needing to specify a spellcasting attribute. This would specifically affect the Magus and martial classes multiclassing into a caster.
When I tried out these changes, here are the benefits I noticed:
Handling saves became a much simpler affair, especially when multiclassing or very specific class features got involved (such as zone effects on Starfinder's Witchwarper). Everything used the same DC.
Multiclassing similarly became much simpler, because there weren't multiple parallel DCs to track. This became even simpler when using the above rules for decoupling spell attacks.
Multiclassing into a spellcasting archetype became generally more beneficial, for martials especially but also for casters opting into an archetype with a different spellcasting attribute. Classes became more generally accurate with spells gained from an archetype, as well as innate spells.
Non-Cleric spellcasters had a generally better time accessing more critical specialization effects. The Magus in particular was a big winner, being able to use the crit spec effect from Arcane Fists with a trained-to-master class DC rather than an untrained DC.
When also using decoupled spell attacks, attack spells ended up feeling significantly better to use, not just because they too were much simpler to use, but because they scaled much more smoothly.
The Magus and Kineticists were major winners with these changes. The Magus using their Strength-or-Dex-based class DC for spells and a physical attribute for spell attacks gave them more leeway for casting spells outside of Spellstrike, though Spellstrike still remained the option you wanted to take when you could due to its action and accuracy compression (Expansive Spellstrike became a fair bit better, too). The Kineticist casting their impulses as spells meant they were no longer isolated from the many mechanics that expect you to be casting a spell, and could more easily opt into stuff like mythic proficiency for their DC using mythic rules.
More specific to Starfinder, but these changes made spellcasting classes slightly more able to "cast gun" using the AoE weapons featured there, notably at early levels.
TL;DR: everything relating to spells became much simpler, these changes made multiclassing, innate spells, and crit spec effects more attractive to certain classes, and the Magus and Kineticist especially had more options available to them without necessarily changing their niche.
Here are the downsides and risks I encountered with these changes:
Because spell DCs and spell attacks were made more uniform, some of those fine-grained differences that came from using different attributes were lost.
Martial classes being able to opt more easily into spells and casters being able to opt more easily into crit spec effects may not be everyone's cup of tea. I don't think this breached niche protection, necessarily, as casters were undoubtedly more accurate with their spells than martial classes, and cases already existed where mental KAS classes like the Thaumaturge or Inventor could get to just a -2 relative to Charisma or Intelligence casters respectively, but it did thin the gap overall.
The Commander and Soldier, while both classes in playtesting, may also be classes to watch out for, as they're martial classes with a legendary class DC. The Battle Harbinger is an odd duck with the intentional split between their legendary class DC and expert spellcasting DC, but the class archetype is just odd and in need of changes in general (from my perspective, at least).
There's this looming fear that the Kineticist casting their impulses as spells means they'll break something. I don't know what that something is, much less have run into it, but the fear's still there.
... and that's my two cents about what worked and what didn't with these house rules. I don't think this is something to implement in-game, necessarily, because I think we're past the point of big systemic rewrites such as this and there are probably a lot of aspects to these changes that may be difficult or risky to implement, but it's certainly something you could try at your table with everyone else's consent if you wanted to. Give it a try if you want, and if this isn't something you'd want at your table, that's okay too!
