Possible examples? Because I still don't see it. What "ranged combat" feature(s) aren't able to impart ranged combat benefits such that the player finds they were actually duped into taking something that fails to perform as advertised? Same goes for melee. Or whatever.
Well, I believe that in the case of the warlock there can be complexities around hand management eg two weapon fighting and a casting focus; or using a two-handed ranged weapon and a focus.
There can also be complexities around, say, not understanding whether your build (in terms perhaps of spell selection, feat choice, CON score etc) is going to be able to maintain spell effects while being hit in combat. To know what counts as a good chance of succeeding on concentration checks, for instance, requires having more than just a casual knowledge of the rules.
Even when it comes to spell selection, it is not always easy to know whether (say) a given cantrip is likely to be effective or not. Consider, say, True Strike. For many people, the choice to take this cantrip will not be so much about "Do I want to tell the story of someone who magically contemplates her enemies defences before striking?" It is more ilke, "I want to magically contemplate my enemies' defences before striking so that I have a better chance to hit them." But knowing whether or not True Strike actually increases your prospects of effectively attacking, especially in the context of relatively at-will attacks, is not a trivial thing: it involves judging the merits of trade-offs between just making two attacks (and thereby getting two rolls anyway) or holding off at range and then closing for an advantage attack later perhaps involving some limited-use enhancer that is wasted if you don't hit.
It can even be as simple as choosing between longsword and shortsword. Does the smaller damage die matter? Probably not if you're fighting classic D&D kobolds (who have d4 hp). Definitely not if you're fighting 4e minions (who have 1 hp). But if you're fighting enemies with large numbers of hp, then you might notice that you are doing systematically less damage than your more heavily-armed friends. These differences can't always be judged a priori - they depend upon getting a feel for the system and what does or doesn't matter. (Eg at a certain point bonuses to hit, or to skill checks, become less important; but to damage tend to keep being helpful as you pile them on, and bonuses to AC in some ways become even better, the more that you have.)
I have seen people have trouble building characters in systems they're not familiar with - particularly D&D, which tends to have many moving parts whose implications and interactions aren't always obvious - and then being stuck with characters who don't perform as they had hoped, because the mechanics don't deliver as had been anticipated.
how can it be said to be a problem with the character creation rules?
I don't know if it's a
problem or not. All I'm saying is that D&D's PC creation rules are among the most complex of any RPG I've ever played, and trying to judge what will or won't be effective in play relative to a given set of desires tends to be far less transparent than in some other systems.
By reputation, at least, 3E is the worst version of D&D for this. The only version that is perhaps immune, because its rules for PC build are so constrained, is Moldvay/Mentzer Basic. Though even then, if you don't really understand the stat rules or the XP rules it can be hard to see what, if any, trade off is involved in choosing Elf over Fighter or MU.
