Does that mean I can run Frodo though the Tomb of Horrors, AKA the best dungeon ever?

Presumably. I see a post saying it has even been converted to 5e in Dungeon 213. At least assuming they publish a stat block for Frodo...

Rules-lawyers seem to conveniently forget how polymorphing works:
"The assumed form can’t have more Hit Dice than your caster level (or the subject’s HD, whichever is lower), to a maximum of 15 HD at 15th level."

"It also gains all extraordinary special attacks possessed by the form but does not gain the extraordinary special qualities possessed by the new form or any supernatural or spell-like abilities."

Of course you can, but I think your halfling rogue is going to need a lot of luck in addition to mithral chain mail and a short sword of goblin / orc detection to survive the Tomb of Horrors... ;)

It seems you only looked at one part of the rules.

First of all, the HD limit isn't a problem because Surrukh have only 12 HD.

Second, it turns out that there are rules that explicitly allow one to get supernatural abilities.

Metamorphic Transfer (http://www.d20srd.org/srd/psionic/psionicFeats.htm): Shapechange (http://www.d20srd.org/srd/spells/shapechange.htm): (Greater Metamorphosis, from the XPH, has the exact same language.)

In other words, that rule may apply when using plain Polymorph, but there are specific exceptions to the rule that can be used to gain access to the forbidden (and game-breaking) ability in question.

Okay, y'all Pun-pun debaters, I'll bite.

The problem in the theoretical build isn't with Polymorph. Polymorph is broken and we know it. The problem is with the Manipulate Form thing, which is open-ended but offers examples of improvements which are piddly shit. Like in every tabletop RPG, if the rules don't say you can't do a thing, you only can if it's in line with other things which are allowed. Manipulate Form, or Wish (Su), or Win the Game (Ex) aren't.

Furthermore, and more importantly.
Theoretical builds are an extremely shitty argument in favor of a game's quality, because you're not going to see one in play. You might however see a monk or a fighter, level adjustments, Spot checks to determine encounter distance, the pile of idiocy which is Hiding rules. These things will spoil your gaming experience. D&D 3e has 666 problems but Pun-pun ain't one.

PunPun is a cool technical optimization, but I'd agree it's utterly irrelevant for a practical discussion of what 3E play actually looks like. You don't need to look at obscure sourcebooks or weird/wonderful technical builds to find problems: the big issues are right there in the core rulebook. The core rulebook is chock full of gamebreaking spells that put the full casters in a completely different league than the other classes. At the same time, half of the warrior-type classes are hopelessly underpowered and require excessive optimization just to get them to function competently. The various combat styles were poorly balance, with two-handed weapons outclassing every other combat style by default.

While I definitely give 4E credit where it's due for improving game balance, that shouldn't be construed as saying it's superior to 3E. While 3E is messier and less balanced, those free-form elements give it something that 4E lacks.

It turns out that the actual Pun Pun setup can be done with the core rulebook using, of course, a full caster. Wizards can do it at level 17. Druids can as well; you just need some way to get a scaled companion (maybe summon a crocodile?). Clerics, who are generally considered overpowered as well, need the Animal domain to get Shapechange, but can still pull this off.

Anyway, I can mention other issues I have with 3e/3.5e that are actually likely to come up in normal play (other than Polymporph/Shapechange issues):

* Character creation takes a while for high level characters. You can't just quickly write up an 18th level character the way you can in 1e/2e (especially if you don't use the optional proficiency rules of 1e/2e). In 1e/2e, you just need a race, class, starting equipment, and starting spells known, and of course rolled stats/HP and you have a usable character. In 3e/3.5e, you need to choose skills and feats for each character.

* The skill system is deeply flawed, especially considering how cheap it is to make items that provide ridiculous boosts to them. This, in turn, makes it nearly impossible to balance anything that uses the system for anything important, such as Epic Spellcasting (easily gamebreaking if you try to optimize it), or the Truenamer (worthless if you don't try to optimize it).

* Too many feats with minor effect. To make matters worse, the feats with obvious, significant effects, tend to be spellcaster feats, and spellcasters, as much as I like them, already have a lot of options. Fighters need feats as powerful as the metamagic and item creation feats.

* Caster level is too important. This tends to make multi-class spellcasters pretty much non-viable, especially with the way multi-classing works. (They actually made some prestige classes to try and patch the issue; the Mystic Theurge being the most obvious example.) I don't like having the Magic Resistance mechanic, especially when saving throws already exist as a defense against spells.

* All spells of the same level have the same save DC. This has the result that save-or-lose spells are far more powerful than things like save for half damage spells. It doesn't help that HP goes much higher in 3rd edition than in 2nd but spell damage doesn't, and that there is a widespread oddly-behaved ability called "evasion" that turns damage spells into save-or-damage but does not affect save-or-lose spells.

* The healing situation has improved from 2nd edition, but there is still the issue that Heal is far stronger than lower level healing spells, and that healing during battle isn't usually a good strategy (before Heal, at any rate).

You do need the Surrukh, though, and that does come from a rather obscure source. PunPun, at his core, is about combining the broken ability of the Surrukh with the broken game mechanic that is polymorph. Everything else in the formula is entirely incidental and meant to facilitate that gamebreaking interaction.

I'd disagree on that one. While there are problems with the way some skills (*cough*diplomacy*cough*) work by RAW, the scaling of the numbers is generally reasonable. Truenaming and Epic Spellcasting are just poorly designed subsystems, and the skill system was a poor fit for what they set out to do.

If I multiclass Fighter/Barbarian, the base attack bonus is additive. If I multiclass Fighter/Wizard then the Wizard is still giving me half BAB progression. Caster level, on the other hand, gets nothing from multiclassing. It really needed to get a similar treatment and have a stacking progression like everything else. That would also mean that non-casters would implicitly have a caster level, which could potentially have had interesting ramifications on the design of magic items.

My only issue with spell resistance is that evocation should have been SR: no. As a game mechanic, having some creatures that are extra resistant to magic is fine, especially given wizards have tools to get around it when it comes up. The bigger issue is that evocation as a school was just nerfed into the ground by all the unfavorable factors stacking against it, and it really didn't need SR as another one of those.

I'd say that's more an issue of the evocation school being underpowered. Save or suck spells are nasty, and many could have been designed better, but I don't feel the DC's themselves are the problem. Rather, the issue is with evocation as a school being underpowered. Even something as simple as making them SR: no would have done wonders for it.

Evasion is a potent ability, but it has limited distribution and isn't particularly common unless you're raiding a thief guild or something like that. Area of effect damage spells are just poor choices against those creatures, and I don't have a problem with some creatures just being highly resilient against certain categories of spells, particularly when the ability isn't all that common. Immunity to Mind-Affecting spells is the bigger issue, since it's absolutely everywhere and it completely shuts down the entire Enchantment school, whereas Evocation still has some spells that don't hit reflex save.

I don't really see an issue with any of that. 3E D&D is a fast-paced offensive game, and I don't see a problem with healing being a poor in-combat strategy. It's still very useful between battles, and the design of the Cleric class makes it easier to have healing powers without having to dedicate yourself to them. I don't see a particular problem with the Heal spell outclassing the Cure spell. Yes, it's a big shift in the power curve, but as far as broken spells go it's not even on the radar compared to the real offenders.

Otherwise I agree with your critiques.

The fact is Pun-Pun is still trying to use class features designed for a progenitor (i,e, God) race. Not to mention how it comes from an obscure setting specific supplement. In short, it abuses the rules and use a very liberal interpretation of those rules. It is not the only interpretation. On top of that, it claims an unlikely "endless loop".

As to multi-classed spell casters, there have been a number of specific Prestige Classes or hybrid classes which mostly addresses those issues. I'll seen very viable Mystic Theurges and Eldritch Knights, even Arcane Tricksters. I'm sure there are many more. Just because any "mutt" multi-class you can come up with isn't as powerful as other classes is meaningless. It should be rather obvious you would be losing something be multi-classing into classes with different prime requisite abilities.

Overall, D&D from 1st through 2nd edition and class kits and then to 3rd edition provides very robust gaming where one can create a world and characters of a vast variety of kinds and types. One can travel through the Planes, play a God campaign, play a game of political intrigue. You can't expect a system that can apply to so much to be perfectly balanced.

That was the mistake of 4th edition. And despite the attempt, they still failed to balance it. On top of the fact it became more of a tabletop miniature game (with bits of plot stuffed in between battles) than actual role-playing.

Tabletop generally refers to RPGs as well, not just board, miniature, dice, and card games.

My apologies for the sudden bump, but I never got around to commenting on this.

This is pleasant news, though not entirely surprising. Thanks to WotC wisely reversing their decision to abandon OGL, we're likely to see more and more D&D compatible products. Plus, 5th Ed went a long way to curtail the magic item economy that was so prevalent in previous editions, making settings like Middle-Earth much more compatible with the format.

That said, this does make me wish Fantasy Flight Games would create a new edition of Midnight that is compatible with 5th Ed, since that too is a very LotR-esque setting.

The DM may not spot the problem right away. It can be touchy to tell a player "You know that ability that you took 7 sessions ago? Yes, the one you build your whole character around. Well it doesn't work anymore."
Esp. with unexperienced GMs or assertive players.

And although a DM can adapt the rules (I tend to mix-and-match and houserule the shit out of my own Fate games), I don't like the "a game isn't broken if you can fix it" argument. No game system can be perfect, especially since every GM and gaming table basically plays a different game (how many GMs did use the full fatigue rules in Runequest? But not using them broke other parts of the game). But a problem that has to be corrected by the GM is still a problem of the game.

Otherwise, we could just say "FATAL rules are okay, since we can ditch them and play with any other ruleset".
Okay, the example may be extreme (and refering to FATAL is the Godwin point of RPG conversations) ;)