1. Emacs and Emacsspeak in particular. Emacs is the richest text-based user experience out there; imagine if all Terminal apps were configurable and scriptable under the same coherent framework. This is why Emacsspeak is an incredibly valuable asset that I would lean heavily on in the event of vision loss.
2. I would invest heavily in making sure that I have a very trivial way to deploy Chrome user extensions or similar such that I could write my own screen readers, document explorers, etc. tailored to my own liking. The only thing close to Emacs is the DOM with JavaScript as it’s Elisp. A lot of people gripe about the web, but the fact that almost all my apps come under the same pliable DOM means I can manipulate without vision almost everything. See e.g. vimperator.
These two things mean that you could do most of your job as-is, with assistance from colleagues.
As you’re a full stack developer, you’ll find point 2 easy. And point 1 is hard but a worthy investment. Emacs is older than the web and will probably outlive it.
I would probably also look at using tree sitter to make screenreading code more efficient at the AST level, because code is parsed linearly but we read it with random access.
I’d start working with my screen turned off or covered with paper if I knew ahead of time, that way I could start training. Honestly, computing-wise I wouldn’t be that worried about losing my vision. It’s the rest of life that’s harder.
Best of luck to you! Drop me a PM if you’d like some pointers on my two bullet points.