I personally don't feel like the shift from 2e to 3e was very jarring, but some of my favorite DMs still prefer 2e over 3e (let alone 4e), because of the inherent complexity in 3e feats and skills. 4e solved a few things but it felt too different. What I'm saying is, going from THACO to D20 was a lot less changing the core feel of the game for me (putting aside feats and skills for a moment), than keeping roughly that same math chassis and overhauling virtually everything else in the game, way too much. It really feels like you can't adjudicate 4e on the fly, and quickly, without seriously nerfing or boosting one or another classes' powers or abilities or feats in the game. E.g. we had our DM in 4e try to limit the number of opportunity attacks you could take in a round to nerf Snarling Wolf Stance because it was, indeed OP. He tied Dex to the number of OAs you could do, not realizing how much this hurt my character who was part defender (a paladin, at that), and needed it, vs the rogue who had like 22 dex and was unaffected. So what I'm saying is, if you make a hard and fast house rule, you could easily break the entire functioning of a class, because he didn't realize at the time the side effects of nerfing one power in a roundabout way were much worse than explicitly limiting that power specifically.
The subtleties of immediate vs opportunity actions were lost on many other players and DMs, of course OAs were always confusing people since the start of 3e, so that's not an edition-specific comment. Simple rules that have good expression in the narrative sense without requiring constant rules lawyering is what I like the most about Next, and in that sense and on that basis alone, it has a good chance of being the best D&D yet. I.e. the D&D ruleset coming into its own after 40 years of ironing out the bugs.
What the players appear to be saying, from what I can tell reading the feedback : Over-Complexity is a bug. Slow combat is a bug. Unrealistic martials are a bug. Too much power for wizards is a bug, as is them running out of spells and bringing out their crossbow. Clerics being OP generally is a bug. Fighters being boring or weak-sauce at higher levels is a bug. Stat / number inflation / arms races for its own sake is a bug. Magic items not feeling special or rare, or being expected in the base math, is a bug. Having godlike stats without magic or supernatural influence is a bug. Having arbitrary duration spells or abilities such as "lasts until the end of the encounter" (when is that, exactly? when the DM says? what if I want my fly spell to continue and not rest?....) is a huge bug.
That said, there's a huge contingent of people for whom powers and AEDU were not bugs, but features. Those were very controversial and IMO don't add to the feeling of playing D&D, thus they should be out. For good. Thankfully the design team appears to be hitting most of the right notes and actually listening to the feedback of the majority of D&D fans, rather than the vocal minority who are a) not game designers, b) hate everything about every edition of D&D except for power cards, or c) just have a bunch of confused and contradictory notions such as making the game only two classes, or removing the d20 entirely. I've seen so much silliness out there I think part of a game designer's job in evaluating feedback is figuring out who the feedback is coming from and what their biases are.
If they spew nothing but venom for the majority of D&D rules's history : IGNORE THEM ENTIRELY. No game designer will tell potential clients to consider playing a different game, but I will.
