You know, I'm tired of debating with people about what they don't like. This whole thread was supposed to be about how to make D&D do certain things better, and it INSTANTLY got sidetracked by people telling us that what we wanted was not to their taste and that they felt compelled to be insulted by the way [MENTION=42582]pemerton[/MENTION] drew a contrast. I'm about just talking about the subject at hand, and I thought that since I basically hacked 4e to do this stuff, here's the essential parts of what I did and why (there are plenty of other things, but they really are not very relevant to this discussion, like altering some of the math).
1. I fixed some of 'the math' so that skill/ability checks and all types of attacks actually follow the same progression in practice, and can thus be mixed together. This is probably the least important change, but it delivers on the promise of true universalism in 4e's mechanics. You can now make a skill check as an attack, it works.
2. I broke check results into graded categories, so you can have total success at a check, or limited success, or you can fail (or even fail really hard, although that distinction is less important in general and probably could be eliminated). This allows a sort of DW-esque "well, you managed to jump the chasm, but now..."
3. Everything, except combat, is a challenge. There are no 'free checks' in my game. You are either engaging in a conflict of some sort and there are stakes, or its 'free play' (interlude) which doesn't need such rules. Checks only relate to conflict as well. I guess a GM could 'play with dice' and use that in his framing, but it isn't part of the rules.
4. Players can buy successes. Using practices a player can acquire a way to buy successes on checks (which you will remember are always part of challenges). There is always a cost, and each practice allows a certain type of fictional element to be introduced into play (IE you might have a ritual which lets you fly, you can use this to create a success, but you still have to explain how it contributes). Players can also simply make the requisite checks and risk failures and thus complications, but avoid the big costs (many practices still have an initiation cost to avoid spamming).
5. Is there a 5? I'm not sure.... Oh, yeah, players can use a sort of inspiration mechanic. It is binary, you can't stockpile inspiration, you have it or you don't. You can acquire it by invoking one of your character's traits in a disadvantageous way (IE I'm forgetful, I misplaced my lock picks). You can invoke it in a positive direction in much the same way, expending your inspiration and invoking a character trait. Players can describe several traits for their character, three usually being enough, and also your other attributes could count, like race, class, etc. when plausible. Generally you can invent some narrative element, achieve a success, avoid a cost, etc. in this way, but only so often. I find it is best to reset everyone to having inspiration at the start of each session, but that could be tweaked. It is a LOT better than the 5e style of this technique!
I think that's about it in basic terms. This makes a much more explicitly narratively driven type of game out of 4e. You could eschew all of this and basically play the same way, but I find that players have a bit easier time when there are explicit guidelines.