I'm afraid nothing with the docs are going to change unless one of two things happen
* Someone with loads of experience with FPC/Lazarus decides to spend loads of time cleaning up the docs, or writing new ones.
* Someone gets paid, as their job, to do it.
The barrier to entry is really high as soon as you try to do anything complex or performance-oriented.
For example, I needed to create a REST API a while back, and had to go with my usual stack of OpenBD CFML instead of FPC because I couldn't find a decent database pooling library.
There are several examples of basic pools and shared resources, but I wasn't looking for something to learn from, I was looking for something to get a job done, something robust.
That's a common story with FPC I'm afraid.
Unless you've already spent lots of time learning the language and learned a good chunk of CS, you're not likely to find what you need that's fit for deployment.
Personally I think it's simply due to the non-corporate nature of support.
There are lots of hobbyists and professionals sharing things, but little in the way of serious man-hour projects.
Take the core phrase "Write once, compile anywhere", in reality most non-Windows targets are incomplete or require workarounds, making the statement just not true.
I like FPC, I like Lazarus, but it's hard justifying using it over CFML, Java, etc. due to the shortcomings for on-the-job deployments and projects.
I give up.
I did not ask for advice for reading.
I asked for the complete cleanup of all existing documentation in such a way that it becomes usable.
But i looks like nobody (wants) to understand.
I will live with this mess as everyone else does here - I won't bring it up anymore.