This is a very good point. But D&Dnext seems a bit ambivalent, at present, about exactly how it wants to handle "story oomph". As can be seen in the current thread about backgrounds and skills, for example.
Which I guess reinforces the point that working out what to do with rangers and paladins is intimately connected to working out exactly how to handle backgrounds and specialties.
If a paladin has more hit dice than other PCs in order to fuel selflessness powers, is it going to be overpowered when used selfishly (I'm thinking of the problems with the 3E cleric who plays as a buffer/controller rather than a healer)?
If I want to play a ranger with the playtest packet with D&D Next, I can pretty much already do it. I I just need a start with a fighter and give him a good dex. Then I need is a background that gives me spot, stealth, survival and nature lore. If I take the sharpshooter speciality, I essentially have the low-level 3.5 ranger. So... why do I need a Ranger class?
Sure, but in 3e/3.5 the abilities which the subclasses (Ranger, Paladin, Bard, Barbarian) had were generally weak in comparison to the abilities that they shared in common with the fighter. What was most important was the shared basic attack bonus and the shared feats (whether they got them as class abiltiies or not). The smite and healing of the paladin, and the spells and skills of the ranger, were both minor abilities that were tacked on a fighter shell.
Specialities can certainly carry the weight of those minor abilties, to the same degree that they are carried in 3.5. If it was good enough then (and still now given the amount who play 3.5) why isn't it good enough until we find something that is more robust and interesting as a mechanic for paladins and rangers?

(Dungeons & Dragons)
Rulebook featuring "high magic" options, including a host of new spells.