Well, I don't know how this compares to 4e casting, because I've never played 4e, and also I wasn't fond of the idea of 4e dumping Vancian magic in the first place. Anything that Mearls is working on comes from the experiences of 4e as well as prior editions, so there's elements I'm not aware of. Mearls is good at analyzing the game and how things work, but I don't always like his approaches to fixing things. It might be because sometime during 3e, the optimization approach started becoming popular and reduced a lot of the game to number-crunching. This is where wizards and CoDzilla were considered the absolute best classes, while everything else was being seen as increasingly worse. That's fine for something like a videogame, but in a tabletop setting, it shouldn't need to dominate, especially if the DM knows what he's doing and gives the non-optimized characters something to do.
Caster dominance doesn't bother me, but since wizard is my favored class, I could be a bit biased. That doesn't matter to me though. The issue here of course is of the quadratic wizard, and this has long been the case. It doesn't bother me because I remember the old school wizard (M-U or mage if you prefer) well. The wizard was a powerful class, but had to earn that power, it was weak at low levels, and had the slowest XP progression in early levels. The eventual payoff had to be earned. The good wizard players knew how to play at low levels, use oil or assist in non-combat ways, do what needed to be done while the fighters were fighting. Of course combat was shorter in the past, monsters had less hp without Con bonuses kicking in, and there weren't things like powers and AoOs and feats and stuff going off. That is when players risked combat, because the big xp payoff was in treasure, and not slain enemies. Then again, the game really only assumed about 10 levels of play while 3e upped it to 20, and really most of the complaints about 3e is how stuff starts to break down in the mid teens. Also, one bonus the fighter gained at high levels was the ability to attract followers. That was in 3e too, but shifted off to the optional Leadership feat, which I'm sure a lot of optimizing number crunchers considered a waste.
At will cantrips sounds a bit over powered at first, but consider that 3e cantrips do only 1d3 damage. That is comparable to the damage wizards could do with what few simple ranged weapons they had in the past. Crossbows do better damage, but the wizard still has to make an attack roll. So this doesn't bother me too much I guess.
Keeping spells under control I think gives a bad example. Sure that cleric might be wearing full plate and have a big penalty to save against
grease, but really at level 15 he should be able to dispel the effect anyway, so why is it a problem?
Dangerous spell casting is a misnomer. This sounds a lot like how it's always been in the past. In Basic and AD&D, taking damage fizzled the spell and it was lost. In 3e it worked like this:
SRD said:
If something interrupts your concentration while you’re casting, you must make a Concentration check or lose the spell. The more distracting the interruption and the higher the level of the spell you are trying to cast, the higher the DC is. If you fail the check, you lose the spell just as if you had cast it to no effect.
Oh look, it's fizzle and loss again, though any wizard who's built at all sensibly will have a decent chance to make the Concentration check. Here the spell fizzles, but it isn't lost, so really it's more generous than the game was in the past.
Don't like the idea behind scrolls. Again, I can't say what things were like in 4e, but if 3e wizards were cranking out too many of them, then why not do something like up the XP cost? I remember that low level scrolls at least had a very trivial XP cost, like 1 XP for a first-level scroll. That's not a huge sacrifice at all even at first level. Wands sounds like it goes back to 2e and earlier wands, which weren't necessarily bad, but again I liked 3e wands. Again if things need control, then make them more expensive so that a wizard isn't just cranking them out at will but must consider the cost.
Buff spells got broken in 3e in part because there seemed to be way more of them than in 1e (can't speak for 2e here). I find people don't tend to use buffs in 1e nearly as much, perhaps because of having to choose between buffing and blasting.
It might also be because monsters have more hp in 3e (again the Con bonuses kicking in), so blasting is considered inefficient. A fireball in 1e and even 2e could clear out whole groups of monsters at once, particularly since 1e did not have a damage cap on it.
Creative use of spells I'm somewhat cautious about, but then I'm also seeing it from an older point of view. I remember how Skip Williams used to advise DMs heavily against this in Sage Advice and the High Level Campaigns book, because creative use of magic could easily lead to abuse.