There's no real big theoretical problems in the quantum computer building space. There's problems of scale, and funding, and usual growing pains of a new industry, but scale went from 7 to 24 fairly quickly and all it took was more money. If I gave IBM $10T dollars, they could build me a 1024-qbit computer. Once it gets cheaper, which is the current problem, I don't see any reason why Azure Quantum (ex) wouldn't simply decrease in price to where it can be used practically.
>There's no real big theoretical problems in the quantum computer building space
The current quantum computers are just on the edge of what we can simulate classically, so we can't yet rule out the possibility that realizing a quantum computation requires an exponential amount of energy in the number of qubits. (Though it should be noted that quantum mechanics predicts that this will not happen.)
This couldn't be further from the truth. If more money could get IBM to more qubits, they would be doing it. Fundamental problems with scaling have held them back for ages. The only people making any progress on scaling is D-Wave and their model is never going to have any impact on factoring anyway.
This is true - in general each bit you add to a quantum computer means the noise floor needs to be twice as low, so difficulty scales exponentially. That is unless a threshold is reached where error correction scales faster.
Only if you somehow "know" quantum computing is ever going to be practically realized. It may never be.