Shell – What does the “opost” stty flag do

Bash needs to put the terminal into character-at-a-time mode while it's waiting for you to type in a command line, so that you can edit the command line using emacs or vi-like editing characters. That's the mode you saw when you looked at the terminal's attributes from another terminal in your example.

Just before it runs a program (in your example, stty), bash puts the terminal back into canonical mode, where you have just a few special editing characters available courtesy of the operating system, such as backspace and Control-W, and basically the program gets input only after you type Enter.

When bash regains control, say after the program finishes or is suspended, it will put the terminal into character-at-a-time mode again.

stty lnext only affects the terminal device line discipline internal editor (the very limited one you get when running applications like cat that don't have their own line editor). For zsh's editor, you'd need to use bindkey (zle does not do like readline (bash's line editor) that queries the tty LD setting to do the same in its own editor).

stty lnext '^Q' start '' -ixon # for tty LD editor
bindkey '^Q' quoted-insert     # for zle

Note that you'd need to do the stty part for every terminal, and do it again any time the tty LD settings are reverted to defaults like after stty sane.

Some systems allow you do change the default tty settings like HPUX with stty lnext '^Q' < /dev/ttyconf.

And for ^V to paste the content of the X11 CLIPBOARD selection at the cursor when in the zsh line editor:

get-clipboard() {
  local clip
  clip=$(xclip -sel c -o 2> /dev/null && echo .) || return
  LBUFFER+=${clip%.}
}
zle -N get-clipboard
bindkey '^V' get-clipboard