Now, have different situations with this step need or possible not need.
Luke, if we not copying .orca folder in /var/lib/gdm folder with target 
system, if any user using for example a different voice speech person 
because using different language, hear different voice the login screen 
output and the first user logged in screen reader output. For example a 
simple example:
When I logged in the live CD, I setted up speech values with following:
Speech system: Speech-dispatcher (this is not changed).
Speech sinthesizer: Espeak
Person: hungaryan
This is need when I would like hear hungarian translated text with 
hungarian Espeak language.

This settings is absolute correct copyed with new created first user 
home directory, but GDM is not sinchronised if we don't do this step.
For example, in GDM, I hear hungarian login screen with the "Cantonese" 
Espeak voice, and this is very horrible. :-):-)
If I or another non english language users would like solving this 
problem, need one time manualy to copy the first user .orca directory 
with /var/lib/gdm folder with following commands:
sudo cp -r .orca /var/lib/gdm
sudo chown -hR gdm /var/lib/gdm
This is not problem me, but why need do this steps with a beginner 
Ubuntu blind user for example he possible absolute not known Terminal 
commands, absolute not known where stored the GDM home directory?
So, I think this is absolute need if we want solwing with non english 
language specific talking problems with an automated way in GDM 
accessible login feature, because this is easy and done now.

The .orca copy in target system/etc/skel folder is a different situation.
For example, I think absolute need provide a basic accessibility support 
with future created users in target system if the system is installed 
with access=v3 option. But, need copy the .orca folder with target 
system/etc/skel folder or not need copying it with normal Ubuntu system? 
This is an optional choice and absolute need you choice. The advantage 
of the copying is a following trivial example:
If a notebook blind user using laptop key binding with the live CD, not 
need always setting all settings if he created a new users. But, I 
understand, what happen if a new user not like the live CD setted speech 
rates, speech pitches?
This is difficulter choice.
Now, Bill and Me working with a pre setted .orca folder in live 
CD/etc/skel folder, because we already setting important orca default 
and optional important settings (Firefox Orca structural navigation 
commands, custom orca scripts, optional key bindings, etc. and we want 
this settings is already setted with new created users home directory. 
Vinux and BeLin using this mehanishm and the users like this method 
since two years. The little problem of we choosed method is of course 
the laptop keyboard layout is not setted correct (we using default 
Desktop layout), but now the patch fix this if need.
In longer period, I think the best way choose is the maximum possible 
integration. This is resulting longer work for example the next or 
next+1 Ubuntu release development, but I would like help and I absolute 
sure Bill would like help same.
For example, now we not setting up extra key bindings with target system 
in 30accessibility script, but in future we easy to do this if have a 
standard keyboard talking map with Vinux and BeLin users. For example 
this is bigger step with next Ubuntu version, but now this is little 
offtopic, but not impossible to do. :-):-)
I would like to do this work with next release.
Sorry my long comment.

Bill, Luke, what the openions?

Attila