Userland and speech-dispatcher issue

[Mgr . Janusz Chmiel]

Dear specialists,
Does somebody of you know, if it is possible to produce fluent speech 
when running Speech-dispatcher with Espeak in user space variant of 
Linux, which support sockets communication and D-BUS?
Is it possible to setup this environment so Espeak will speak fluent? 
Without very often speech interruption?
Or such non standard environment of Linux can not be supported by 
Speech-dispatcher?
I will tolerate one second delay between key press or spd-say command 
and Espeak speech.
But I Am sad if speech is very interrupted.
I Am ready to follow all required steps to debug according to yours advices.

Thank you very much for yours support.
With kindness regards.
Janusz Chmiel