My guess is that "simulation" is probably a good catch-all for why 4E does not fit the bill. It has too much carefully integrated complexity that has zip-all to do with anything beyond its self-referents. That doesn't make it a "bad game" any more than similarly abstract qualities make Chess bad; that may, however, make it an unsatisfying substitute for a traditional role-playing game.
The point about OD&D seems to make clear that the OP wants more quantification of non-combat activities. That suggests a couple of questions:
(A) How much should capabilities be "game currency?" Hero, GURPS and 3E put a lot of emphasis on the "build," which seems pretty typical of modern games. You've got only so much of a (game-contrived) resource to "spend" on abilities.
In RuneQuest and other Chaosium games, by contrast, a character can have any number of skill ratings at any levels; the constraints are "in-world" factors (access to a teacher, opportunity to practice, rate of learning from use in the field, etc.). One consequence is that one can expand or contract the list of abilities rated, and adjust the ratings themselves, without much worry about "breaking a system."
On the other hand, some people find the absence of such a "balancing" system a lack; manipulating it is a "sub-game" they enjoy in itself.
(B) How extensive should the detailed treatment of specific situations be? RQ and (IIRC) Rolemaster and Palladium are examples of games that largely leave it up to the GM to decide what sort of bonuses and penalties should apply in a given situation.
Some other games devote quite a bit of space to defining tasks of various sorts. What's the difference in difficulty between jumping 6' and jumping 8'? With and without a running start, or while wearing a mail shirt, or on muddy ground, or after a day of hiking on a cool day versus a hot one... and so on.
