I can understand the argument that making software more accessible and usable is bad, especially given that the mdern attitude towards software seems to be one where everything should "just work by magic", no knowledge required. This is a deeply unhealthy attitude to have, as it renders the entire concept of learning and exploration meaningless or, even worse, a source of frustration.
I simply disagree with categorizing over-simplification as accessibility as, more often than not, oversimplification simply strips a tool of its utility (usually debugging), and thus is less accessible in spite of the attempts at making it more accessible. A tool that provides terse debugging is more accessible to the end user than one who hides it, this is because software is prone to error (including user error) and communicating said error to the end user is far more accessible than needing them to read the source code and triage the error themselves to fix it. This is just software engineers and designers being bad at their job and being bad software.
That said, I can understand the confusion on definitions here, hence why you replied.
Anyway, to stay on-topic, it's imperative that we improve tooling in a way that encourages LEARNING
Learning a tool is utterly necessary for accessibility of said tool. If a tool is impossible to actually learn, then the user will never be able to actually use the tool to its fullest potential or know when the proper time to use the tool is.
The anti-design of oversimplification of a tool, even if it is to lower the initial barrier to entry, is always anti-accessibility by nature. A good tool can make it easier to pick up for a beginner and capable of nearly anything in its targeted problem space for a master. I think FFmpeg is a great example of a tool like this, its ffmpeg -i video.mp4 video.webm to change file type, but a master of FFmpeg can write insane scripts to do all kinds of crazy stuff with the tool.
That's what makes a big difference, in my eyes, between actually simple software and over-simplified software: Does the simplicity detract from overall functionality or the user's ability to learn? If it does, then in the end accessibility is hurt. If not, then it's just actually well made.
but unfortunately I think the current generation of computer users are too far gone and too stupid to actually learn anything about technology, so maybe the war is already lost.
This is more of a pervasive cultural problem around computers than anything, mostly because there is no generational wisdom to their use yet as Gen Z has not had kids.
Take cars for example, here in the USA, it's the cultural norm that a parent will educate their child on a car, its functionality, and potentially even its maintenance and usage. Because of this, there is a higher-than-normal level of car competence that the general population has.
Even rather goyish people who just outsource all their problems with money ultimately still need to know how to drive the car, lending to some level of competency (and potentially a pathway towards further competency if they ever decide to learn something).
This culture needs to be exported to computers. I figure in due time, this will happen as the more well put together millennials and members of Gen Z that are capable of having children will teach their children these tools.
Because of that, people who have technical competency should do their best to make software that is accessible, but not over-simplified, so that non-technical people can easily learn the tools and pass that along. If enough people are capable of competently using computers instead of being goy about them, then the tides may be able to turn.
I especially believe this could be the case if tech giants start careening and failing due to their own blind drive for profit (Microsoft's obsession with abusing AI to rewrite much of Windows comes to mind as a sign of sickness). So long as freer computing mechanisms are able to be taught to normal people we will be fine, unfortunately though, I do not think there are enough people working towards documentation and accessibility to make traditional FOSS and Linux viable long-term.
