Yes, there will be some hybrid combos that work poorly, but it should be rather apparent from the outset that a given combo won't work well.
I mean, if you want to play a suboptimal combination, fine. I don't think that the poor choices from an optimization standpoint should set the bar, though. You should look at things like thaneborn barbarian/inspiring warlord, or wizard/swordmage - combos that at least share main ability scores.
I will say, though, that it will likely be hard to make truly effective hybrids, unless both classes fill the same role and share a main ability. Getting half of each of two roles' role-fulfilling abilities can really cripple you, especially in 4e where the idea seems to be to have specialized characters come together to make a well-rounded team.