As far as I can make out it is a question of what accuracy you want. For mm
accuracy I'm sure Dr Hill's comment holds, but if you just want cm/m
accuracy then a static/base unit in the locale of the area of interest would
be sufficient (it is likely to be using the same satellites anyway). I was
using a Leika DGPS system just this weekend and the correction signal was
beamed from a site in Wales (Holyhead ->I was mapping in the centre of
Ireland ~250miles away!). The repeatability of the resultant map was quite
good - enough to show that the mappers did take the same path when doing the
work. Of course when the differential signal was lost the plot jumped west
by about 5m.
I'm working on a system similar to your idea but because of the terrain I
will have a centrally located unit, static and just logging position (this
should wander with the dither). The second (mobile)unit for the mapping I
will carry with me and at the end of the day I'll subtract one data set from
the other to get (hopefully) a more accurate map (and hopefully zeroed on
the static location to boot). I only need 1~2m accuracy so I'm hoping that
will do.
I'd be interest to hear your - or anybody else's comments on my proposal.

John

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics