Wit this text:
> Play with it a few times you will see my point. The command you need is:
>
> $ echo hello
When I use Ctrl+q at the 1st line, I get this
> Play with it a few times you will see my point. The
> command you need is: $ echo hello
Where what I need is:
> Play with it a few times you will see my point. The
> command you need is:
>
> $ echo hello
In short, emacs seems not able to tell where the paragraph ends if it is quoted.
If I understood correctly, you are looking for the command message-fill-yanked-message, which is bound to C-c C-q in message-mode by default. See the Message Manual node on Insertion for a brief description of this command, including a caveat.
In fact, message-mode is so powerful and flexible that I recommend reading its whole manual, which is relatively short, and/or reading through the user options it provides in M-x customize-group RET message RET.
Note that message-fill-yanked-message obeys the buffer-local variable fill-column and fills every paragraph in the message. For filling a single paragraph you can use the command message-fill-paragraph, which is not bound to any keys by default. When given a prefix argument, the latter command also justifies the paragraph.
message-fill-yanked-message is not only bound in message-mode, it only exists in message-mode. You get "No Match" if you attempt it in foundation mode.
Commented
Apr 16, 2019 at 0:56
message-fill-yanked-message does not exist solely in message-mode. The function is defined as soon as message.el is loaded, and will continue to exist, i.e. be defined, until you unload the library.
fundamental-mode is intended as a bare-bones mode without any features. Specialised editing modes like text-mode or message-mode depend on the setup that is performed in the eponymous mode function, so their specialised features, such as message-fill-yanked-message, will often not work without this setup. It is wrong to assume message-fill-yanked-message will work as intended outside of message-mode buffers, unless the documentation explicitly says so, or you discover through experimentation otherwise.