The real roadblock might be that there are limits to how many bits there can be in a register. NIST's recent Quantum Computation conference included discussion of the expected lifetime of a computation (what fraction of a second the computation would have to complete in before the internal state space goes incoherent). The more bits are bound together, the shorter the lifetime of those bits, according to one result. However, the more bits you have the longer the computation has to be. This suggests that any given Quantum Computer technology point will lead to a maximum state size (likely in a small number of bits) for a given application.
Something that might be relevant here is that relationship between energy and lifetime for virtual particle generation. When a virtual particle is generated it can have a random amount of energy. However, the larger the energy level is the shorter its lifetime is. It is related to Plank's Constant in a relationship that I don't have on hand, but should be in most quantum texts.