Part of the issue is that versatility in much less powerful in 4e than in 3e. In 3e, many monsters were completely immune to whole swaths of attacks, or might be terribly difficult to defeat by some means but easy by another. Using the right tool for the job was very important. Some non-combat encounters were dominated by magical effects that you either had or didn't have (say, water breathing or what-not). Because raw power often didn't matter nearly as much as just having the right tool, sacrificing focus was often a good choice. That, combined with prestige classes and multiclassing rules that commonly made it possible to retain the core of your class (at most giving up some small amount of power) while gaining new, different options, meant that good characters commonly were multiclassed or prestige-classed.
In 4e, a small to-hit bonus actually matters. Resistances and Vulnerabilities are rarely absolute; in fact, if you can deal just a bit more damage than your neighbor, even if you're attacking something that's resisted, you may well end up doing better. Raw power is very important, and flexibility (though still valuable) less so. Out of combat most classes get only few relevant features in the first place, and rituals (which can be expensive) are one of the few remaining options.
So when in 4e a hybrid rogue gives up sneak attack partially, this loss of power really matters. When a hybrid paladin loses much of his marks power, and need to use more actions to boot, this really matters. Having a high-attack stat matters, but for hybrids this is usually hard to achieve, since it's unlikely the stats actually line up perfectly. When a hybrid loses armor proficiencies, it matters.
For all these reasons, a hybrid loses out on the power scale. It's possible to make hybrids that work, but it's hard. The only characters that have a chance of competing will need to have just 2 relevant stats, or need to have just one attack stat and 2 secondary stats that don't need to be very high (i.e. neither may be the AC stat for a light armor character, for starters).
Many hybrid features limit their usability such that they can only be used partially, in conjunction with that classes powers. But that means that at any given moment, a hybrid character tends to benefit from far fewer abilities than a single-class character (since hybrids also tend to be missing various extra's).
Even losing just +1 to +2 attack and damage versus a single-classed character is going to be hard to compensate for a hybrid character, and those areas where features might synergize tend to be carefully compartmentalized by the rules. The only real win tends to be the greater power selection pool; and that's worth something, but I doubt that it's worth even +1 to attack and damage for most characters, and it's almost certainly not going to be worth all the various smaller supporting features you lose, the armor proficiencies, and the poorer stat distribution most hybrids will face.
So, if you can find a hybrid combo with a decent stat distribution, where the "extra" features aren't all that important (say a wizard) and where the armor proficiencies and weapon/implement usage isn't too problematic (i.e. where you're basically giving up very little compared to a single classed character) then you can make a fine hybrid - perhaps even an overpowered one if you manage to find some feature combo.
But it's unlikely - it's much more likely you'll end up with a much weaker character with fewer features that work only more situationally, with a broader selection of powers to choose from.
The problem is, most classes already have a very nice power selection, and the extra options tend to be pointless unless they're broken - so now we either have weak hybrids, or abusive overpowered hybrids, but not much in between.
I don't think it's a huge problem; it's not quite as bad as paragon multiclassing, but I do expect only very few hybrids to be made.
