I agree that the applications for drugs design and material simulations would be the near future for QC, but unfortunately my professor seems not to focus on them. That's another source of my anxiety: when other groups are considering Fault tolerant frameworks, we are still at the most physical layer, and left behind by routes like neural atoms. So that may be more likely a personal trouble lol.

As for the sensitivity of qubits, they're much more sensitive than integrated circuits and it's far more expensive to keep a suitable environment for the former. And there are natural limits for decoherence.

Thank you for your insightful comment!