It's rare that the first page of a book will tell you everything you need to know about it. But the introduction to The Epic Level Handbook
(Andy Collins and Bruce R. Cordell) wastes no time in revealing the book's greatest weakness. "The difference between epic magic items and artifacts is that artifacts are unique items generated by a one-of-a-kind event or forging. Many epic magic items are just as powerful as artifacts, but epic characters know how to make them."
This profoundly misses out on what makes epic tier play appealing, and it kind of sets the tone of the whole book. Artifacts are magic items with a deep connection to a setting's lore, created by powerful beings who are more myth than history, and at no point in the entire text does it seem to realize that maybe that's a niche for characters who have transcended the core book's standard level 1-20 progression.
There a planar metropolis, called Union, that acts as the centerpiece of an epic level mini-campaign. The description of its police force gives the stats for a typical guard patrol - two level 14 fighters accompanied by a level 23 sergeant. And if those guys can't do the job, they can call in a back up team . . . 2-5 level 31 characters!
That's the sort of game this book expects you to play. You're level 21+, but you still have to worry about city guards. Never mind conquering the universe, you can't risk being caught shoplifting.
And the level 31 town guards aren't a thing in the setting, either. You might think it could be played for parody, but it isn't. Chapter One gives a list of characters that represent the benchmark for epic level: "Baba Yaga. Conan the Barbarian. Cu Chulainn. Elminster of Shadowdale. Elric of Melnibone. Fafhrd and the Gray Mouser. Gandalf. Gilgamesh. Hiawatha. Odysseus."
Now, aside from doing a terrible injustice to levels 13-18, imagine picking 1d4 + 1 people from that list and paying them 155 gp per day (the canonical wage of a level 31 Union Sentinel) to patrol the marketplace and apprehend pickpockets. I mean, in a multiverse setting I suppose you could have an entire city that was populated by nothing but people who would be culture heroes in their worlds of origin, but something like that would be a lot weirder than the fairly conventional trading town presented in chapter 6. The closest thing I've ever seen in fiction is probably the Citadel of Ricks from Rick and Morty, and that had two whole episodes that went into detail about how unworkable the concept would be.
But that's this book all over. Sometimes, it gets close to an epic feel, but mostly it's committed to telling the same types of stories you had at levels 5-9, but with bigger numbers and with martials being even more of a sidekick to the casters.
Oh, fuck, now I'm remembering it. The Epic Level Handbook has a good idea - allow characters with high skill levels the ability to transcend the possible and perform feats out of legend. If you roll well enough with the Tumble skill, you can safely fall from any distance!
Perfect. Except that "well enough" in this case is DC 100. Just a reminder, the maximum skill rank is your level +3, so that's level 97. Except that you'll probably have some ability to take 10 on Tumble checks, so that's 87. And then maybe Epic Skill Focus, for an extra +10, so 77. Also, attribute bonuses, and other magic item bonuses. Let's say it's an even level 50. To do a feat you could replicate with a level 1 spell!
Oh, I suppose a level 1 wizard could only shorten their fall by 60 feet before Feather Fall wore off, and if I'm extrapolating correctly from the chart (which shortens a fall by 20 feet at DC 30, 30 feet at DC 45, and 40 feet at DC 60), then that would only be a DC 90 check. But a level 2 wizard doubles that to 120 feet. A level 20 wizard gets you approximately a quarter mile (assuming you wouldn't just cast fly or polymorph in that situation), So, I guess there are situations when having inhumanly perfect mastery of a skill might beat out low level magic, but maybe before I make any judgements, I should double-check the new epic spellcasting system and see what a Spellcraft roll can do at a similar DC . . .
Ah, here it is, DC 97 to cast Damnation, a spell where you "send your foe to hell."
Okay, maybe that isn't the best example because it's an inefficient use of the spell creation system (for example, with a similar DC, you could create an instant death spell with a +25 save DC), but the contrast is palpable. Casters get much nicer toys than non-casters.
Not just with skills. Feats as well. Buy a feat that reduces the level adjustment of all your metamagic feats by one or a feat that increases your weapon focus bonus from +1 to +2.
Some of this I have to attribute to inexperience with the 3rd edition rules. When characters advance past level 20, they continue to increase the strength of their class feature and get bonus epic feats. And so, the rate at which you get epic feats depends entirely on how many class features you have. Fighters continue to get a bonus feat at every even level, because that was their level 1-20 pattern. Barbarians get one bonus feat every 4 levels, because their uncanny dodge and damage reduction continue to improve (one point every three levels, staggered so that you only get one bonus at a time). Now, keep in mind, Damage reduction is an epic feat that gives you 3 extra points of damage reduction. And it's on the fighter bonus list. So if both the fighter and the barbarian were investing every available resource into damage reduction, then the fighter would have better damage reduction at level 24 (DR 10/- vs DR 12/-). Which isn't to say the Fighter is overpowered or anything. The fact is, the gap only exists because of the rapidly diminishing returns from advancing low-level abilities (what's one extra point of damage reduction, that only works against physical attacks, going to do for you at level 29? what are you going to do with 8 rages per day that you couldn't do with 7?), and the non-magic feats merely have a slightly slower rate of diminishing returns (because the three points of DR you get from the feat is scarcely better at these levels).
What makes it an absolute shitshow is that mage feats do not have these same diminishing returns. The Automatic Quicken Spell feat applies the core book's most expensive metamagic to all of your level 0-3 spells. Buy it two more times and you'll Quicken all spells on your standard advancement list, even ones that are too high level to cast with metamagic (you'd have to buy 4 levels of improved spell capacity to get the 13th level slot to Quicken a level 9 spell, and you could only do it once per day).
It's probably not a coincidence that every single one of the Forgotten Realms epic characters had multiple caster levels (and all but one was in the double-digits). The meme of "Linear Fighter vs Quadratic Wizard" predicts a widening gap between casters and non-casters, and this book is determined to prove the meme correct.
I guess, overall, I just found this book to be incredibly disappointing. This was an opportunity for the game to really cut loose and indulge in the impossible, but it kept wanting to wrestle the characters into the same old roles. I haven't even addressed the monster design, where they pitch a perfectly fine level 10 creature like the brachyurus, which is essentially a bigger, meaner dire wolf, and then inexplicably give it 38 hit dice. Yes, I suppose this mostly mundane beast is a CR 23 encounter, given that you made the numbers so much bigger, but it's still only a Large creature, same as an ogre or a horse. I refuse to believe that this is a more legendary fight than going up against the Terrasque.
It's funny, but I don't remember being so upset about this book in 2002. I must have had the ability to notice its terrible math (possibly the worst I've seen, only Initiates of the Art comes close) because this was the same year I was acing combinatorics (yeah, it's a brag) and I've only gotten worse at math since then, but in those days, I don't think I scrutinized my games so closely. I never asked myself how many levels my characters would need to deliver on the promise of the back cover, or what it would be like to run every class in the same party. I think it's because I was deep into Aberrant at this time (talking about terrible math) and so I wasn't really invested in having D&D that could work at this power level. I think I just bought the book because I approved of its concept. Maybe I'm lucky that I never got to use it in a real game.
Ukss Contribution: There's only one thing that could go in this space. An idea so distinct and memorable that it stuck with me for 20 years, even though I've only read this book once or twice. I knew, even as I was picking the book off the shelf - "the Ukss contribution is going to be the atropal." It's super edgy, but also one of the few things in the book that felt like it belonged in an epic story and not just a high level campaign. They're the undead fetus of a stillborn god. The idea is creepy. The art is creepy. And the description is creepy: "It constantly drools stinking ichor as it mouths obscenities."
For all of my frequent complaining that I'm not a horror guy, I've got to recognize when someone does horror right.