Hi Petr,
I just saw your post. Do you work only in the Wolfram Cloud or do you also use the Wolfram Desktop version? Things like keyboard navigation work better in the desktop version.
I also recently made a function that can read out the output by using AI/LLM functionality:
https://resources.wolframcloud.com/FunctionRepository/resources/CaptionedEvaluate/
I don't know if this is useful to you, but I am curious if it is.