3.0's glass cannon outsider spell like ability platforms versus the 3.5 beefed up combat stat bruiser ones who can fight better at their CR class but have stripped down magic.
The funny thing is that even in 3.5, the creatures with big lists of spell-like abilities (SLAs) were quite often not the combat monsters their CR presented them as being.
Sort of.
If you ever picked up a copy of Bad Axe Games'
Trailblazer: Teratologue (affiliate link) – which was a monster book companion to
Trailblazer, their own attempt at a 3.75 edition – you likely noticed that they made an interesting notation regarding the monsters' CR. Specifically, it noted that they'd used the work of Craig Cochrane (our very own
@Upper_Krust), who had taken the idea that a monster's CR was the aggregate total of everything in its stat block to its natural conclusion, devising an intricate breakdown that assigned a numerical value to each aspect of a monster's stats (i.e. Hit Dice, Base Attack Bonus, feats, skill bonuses, movement types, etc.), all of which cumulatively added up to its CR.
Now, the
Teratologue didn't reproduce U_K's work in full, but instead adjusted the monsters' CRs accordingly. However, it included a certain takeaway from it: all of the monsters in the book had their total CR, but also a "spine CR" listed.
The "spine CR" was the Challenge Rating that a monster had by taking into account only the "spine" of its stat block, which was the term that Bad Axe Games used for the core features such as Hit Dice, Base Attack Bonus, melee and ranged attacks, saving throws, and one or two other things. The idea there was that the spine CR was an accurate measurement for how dangerous a creature was if you treated it like a brawler; that it ran up to the party and just started making attacks for all it was worth.
For plenty of monsters, the spine CR and the total CR were virtually identical. A bulette or a dire bear aren't going to do much besides move up and make attacks each round. But for creatures with large amounts of SLAs (or actual spells, for that matter) the differences between the spine CR and the total CR were often dramatic. The pit fiend (which was virtually identical to the 3.5 pit fiend), with a total CR of 20, had a spine CR of...9.75. Which is to say, by having it plant its feet and slug it out with a 20th-level fighter, the pit fiend was going to absolutely get its ass kicked in short order.
Total CR takes
everything into account, and for creatures with large lists of SLAs, those make up a considerable portion of that number. So if they act like casters first and foremost, before eventually closing to melee, you get the full "value" of the encounter. But if you have them move in and ignore their SLAs, then it shouldn't be surprising that they're not going to have as much oomph as their total CR would suggest.
It really made for fascinating reading, and it's a shame that the concept (and the book) isn't wider known today.