Doing simple arithmetic quickly and with high fidelity was instantly useful and valuable. There were many other early use cases for digital electronic computers that weren't valuable. You don't know about them because, since they couldn't turn a profit, they don't exist right now. You should read up on early Soviet computers and Soviet cybernetic theory to get a sense of other things that were possible, though ultimately not profitable.
My issue with quantum computing is that there's no real evidence it can even do the things it's been promised to do. Did you know there aren't even any proposed quantum algorithms that could factor large primes faster than conventional computers? Nevermind the number of qubits you would need to accomplish them. There's no theoretical framework for how they could even do that. It's entirely hypothetical at this point. To create an algorithm that can reliably output useful results while maintaining the coherence of the quantum states has proven to be significantly harder in practice than earlier hypothesized.
Much like quantum encryption, it turns out creating the system in theory is much different than making a practical system. It's a perfect example of a spherical cow problem, also called an AM/FM problem. While your theory of how to increase milk productioni works for spherical cows in a vacuum, it doesn't work with real cows. It's the difference between Actual Machines and Fucking Magic.