Also there is a much easier fix to this whole problem if you still want to play 3E, but don't want the headache I presented. Mournblade mentioned spell interuption. Bring that back into the game. Don't allow concentration checks when a wizard is hit mid spell. This is a minor change that can balance things out some more, if you find the concentration checks are too easy to pass.
I don't like the idea of making magic (and remember we're talking about more than just a single character class here) even more swingy. The solution you propose is really the same sort of tool 3E already uses to control spellcasters with spell resistance, high monster saving throws, and such. "Your spell could do something that would alter the entire encounter if it worked, but the odds are stacked against it working."
It's actually weird to compare to fighter types. At low levels, fighter types miss a lot of the time but rarely have any trouble with the getting into positon to attack or making their damage significant part. Meanwhile spells usually succeed. At higher levels, fighters are nearly guaranteed to hit most opponents but have to work harder to make hitting matter. Meanwhile spellcasters "hit" somewhat less.
Just doesn't appeal to me; as 3E doesn't appeal to you.
You have this funny idea, which you state several times, that 3E doesn't appeal to me. That I don't like 3E. In fact, I like it a lot. It's an elegant system with a lot of appeal, and I've had a lot of fun with it.
Many of its features, including features that 4E dumped, have a lot of value to them. I really like the idea of a unified system of mechanics for PCs and monsters, and I would like it without having to pay too high a cost for it. Third edition made a lot of very difficult to implement mechanics work... some of the time.
Maybe I even like 3E better than you do, because I like it enough to really acknowledge faults I've seen in it. Some things that 3E does, especially connected with magic and the skill system, simply start to break down at high levels. And the thing is, a lot of what you've been writing is perfectly true. It is possible to address those problems with tools already existing in 3E. They don't necessarily lead to a bad game experience.
It starts to become a matter of cost/benefit more than anything else. Are the features that become problematic worth the effort the system has to go to in order to preserve them? Are they really that valuable?
You don't have to. If you don't like magic that way, play another game. Some people actually like using the tools 3E offers. Its a toolbox system. I like the magic in 3E. It is cool, it is a great plot device for GMs, and it adds a sense of wonder and suprise that really makes me enjoy the game. Sure wizards are more powerful at higher levels. I accept that. If that creates problems, though, there are ways to manage it.
See, I think it would be possible to get that sense of "wonder and surprise" and use those plot deviced while still dumping all the problematic stuff. Heck, in late third edition that's the way I saw the game going even within the 3E framework.
Take the Eberron setting which both made magic supremely controllable and unwonderous on the PC level while at the same time encouraging the DM to use 'eldritch machine' or 'draconic prophecy' plot device magic that obeyed no rules but the rules of narrative and story.
To me, that's where the wonder in magic comes from. The stuff a character tosses around in a fight isn't really magic. It's just a different form of boom.