Your questions are helpful bar-setter for me, and more or less align with the questions that I had when I was starting out this project (sans the skepticism of quantum mechanics period, I take that as a given). Going down your list:
- Yeah there's a distinction between asymmetric and symmetric encryption schemes. Asymmetric schemes are typically used to make a shared private key which is then used in ensuing symmetrically encoded communications. Those asymmetric schemes are broadly vulnerable to quantum based attacks, hence the need to upgrade to 'post quantum encryption schemes' (PQS). PQS approaches have been developed and are slowly being rolled out, even though it's unclear when the threat of quantum-enabled cracking will be real.
- Yes, I cover this extensively. This actually relates to your last question as well, since error depends in part on what kind of qubit platform you're working with. A superconducting qubit naturally 'decoheres' (loses its unique state) over time, with some sort of semipredictable rate of decoherence, whereas photonic qubits sometimes just get lost! All platforms have some sort of built in error due to the fact that you are applying essentially analog gates to them, and these gates have some imprecision that may build up over millions of operations. I'd characterize the challenges as A) reducing error, and B) correcting the errors that inevitably occur.
- This was one of my sticking points too. The short answer is that there are a few different modalities all competing to be 'the one', and no one really knows what's going to win out. They all have their own (dis)advantages.
The latter points were things I gathered from skimming recent headlines and articles. I should read more thoroughly.