I read many different documents, web pages, source code etc every day. Emacspeak and/or emacs provide all the functions necessary to read a document by line, by paragraph, by page, beginning of buffer to cursor, cursor to end of buffer, interactive mode (i.e. read paragraph and wait until you hit space to read next paragraph), by region, by sexp. Likewise, speechd.el provides the basic functions, though lacks some of the more high level functionality of emacspeak - but this is emacs, so it is fairly trivial to create custom functions to do whatever interaction with buffers you want.
I work as a developer and have to interact with remote and local systems on a daily basis. There is also a fair amount of sys admin and dba work involved in what I do. Eamcspeak is able to meet all my requirements and a primary reason why I wanted to try and understand why you found it was unable to do what you needed, requiring another package which seems to reproduce something which is already available.
neither emacspeak or speechd-el offers me the possibility to read a text
and move throwgh it while in reading.
I have the idea that you think i don't know anothing about emacspeak and
highlights what I'm trying to understand as both emacspeak and speechd will do this in different ways. For example, emacspeak-speak-browse-buffer will speak some of the buffer and then wait until you hit space, then moves cursor to the next 'chunk' and speaks that etc.
From your last post, things are a little clearer. I can appreciate the idea of creating a package which only sends the content of a buffer to some back-end speech service. However, I think you will find there is still a lot of complexity and challenges there if you want to create something robust enough it will be useful to others. This is why I suggest using speech-dispatcher on the back-end. Emacspeak does not use speech-dispatcher and I would estimate that the most common support issues with the package, especially for new users, is getting the speech servers to work. While things have got better in the last few years, the different Linux distributions, different TTS engines, different package versions etc, make this complex and difficult to manage.
One of the first things you will need to do is move your source repo to a 'free' (GNU approved) repository hosting service - github.com