I started doing exactly this. Write a bunch of functions, that may end up in different systems, on different machines, even. This allows you to define the interfaces, figure out data dependencies, and so on.
The code may be runnable, just printing out some statements. Then I copy blocks of it to the place where it will belong.
It’s more of a thinking tool, than “actual code”.
I have set up forgejo, which is a fork of gitea. It’s a git forge, but its ticketing system is quite good.
I know you asked about VMs, but fwiw there are GPU-capable containers now: https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html
Used one of these and the setup is as easy as it sounds. It can run Houdini, Stable Diffusion.
Concepts like Reactive programming are widely used in web/UI contexts. The problem of connecting a UI to an underlying data set is not trivial. Several frameworks deal with this.
As was already said, concerns like Accessibility are studied academically. They have more to do with user experience than the technology, so not sure if they match your question.
Same. Really happy with it.