Also things like Vancian magic ARE the same from OD&D to 3.5 rules-wise. It is the CONTEXT that makes it work differently. OD&D magic users worked pretty well. They could be powerful, but in the context of the dungeon crawl/hex crawl exploration game of OD&D, and with limited size spell lists etc they weren't really that bad. The same basic rules in the context of 3.5 are totally borked. I CAN group those games together in some respects because they share mechanics. OTOH when you play them you'd never mistake one for the other, and if you could just play OD&D there'd be little reason to change the magic system.
Context changes were important, but much of what breaks the 3.5 wizard compared to the 1e M-U isn't context changes, but mechanical ones. Off the top of my head...
1) 1e M-U's couldn't cast in combat period; 3e Wizards have relatively easy means of casting reliably in combat (5' step, combat casting)
2) 1e M-U's if struck while casting a spell had it automatically disrupted; 3e Wizards have relatively easy means of casting spells through hardship (concentration checks)
3) 1e M-U's faced the problem that as their level increased, targets would be increasingly immune to 'save or suck' (pass saves) because saving throws always trended 'better'; 3e Wizards can expect as their level goes up that targets will often become more succeptible to 'save or suck' via there increasingly high DC's.
4) 1e M-U's faced the problem that their hit points were basically hard capped at 10 HD, that consitution was difficult to raise, and that even at 18 Con they were hard capped at a maximum of +2 hit points per level, and that they died at 0 hit points. This meant that most M-U's never had more hit points than the mid 20's, even at quite high level and as such were always one bad round from death. But 3e Wizards aren't capped on their HD, find constitution easy to ammend through standard belts, gain the same bonus hit points as fighters, and most Wizards can actually expect more of thier hit points to come through con bonus than they do through their HD. Indeed, the high expectations about Con bonus mean that low base HD is relatively less important. A wizard no longer expects to have less than half as much hit points as a fighter of the same level. As such, one of the wizards main mechanical drawbacks is greatly mitigated.
5) As a minor point, 3e Wizards saw their BAB go up at a 1:2 rate and they gained multiple attacks per round; 1e M-US saw their BAB go up at a 2:5 rate, and they never gained multiple attacks per round.
6) A great many powerful spells which had severe restrictions in 1e (haste aged you 2 years each time you cast it, polymorph other forced the target to save or die and frequently altered the mindstate of the target to the shape it found itself in, polymorphy self didn't even let you make attacks, etc.) lost those restrictions in 1e.
7) 1e M-U's didn't get bonus spells like a cleric; 3e Wizards did, additionally, many 1st level spells shifted down to being 0th level spells and the ultimate result was effectively many more spell slots than an equivalent 1e M-U.
8) 1e rounds were divided in to periods called segments, and casting a spell occurred over a that period of segments during which it was vunerable to disruption or by which time it might have lost a certain degree of relevancy. By contrast, in 3e each persons entire turn plays out in a single segment of a round (which in turn has many more segments), which means the only danger of disruption beyond easily avoidable AoO is held actions - a mechanic that forces you to forgo actions in the action economy to use and so is rare, clunky, and often inefficient. Therefore disrupted spells are very rare in most campaigns.
There were a lot of context changes as well, particularly in what we might call the 'default' game world. But I don't believe those were the most critical changes. The details of the implementation are important. In a Vancian system, balance is determined by the implementation of the spells individually. It doesn't matter how balanced the system as a whole is, if the individual spells - the little packets of rules they represent - aren't balanced then the class that consumes them isn't balanced. Much of the balance in 3e can be brought back with some fairly minor rules adjustments and spell revisions. I await high level to truly prove that, but so far things are working out perfectly and I have one more tweak to play (bringing back casting time) if the melee/martial types start losing spotlight that I'm pretty sure will close any balance gaps that begin to develop (if they do) around 11th level or so.
