This isn't right. The acceleration in O'Neill's mass driver was much
higher that that. Heck, I think the 3rd prototype at SSI hit 200+
gees. It was also designed to launch payloads much smaller than
a person (a few kilograms at a time, at a high rep rate, with the
buckets being decelerated and recycled.)
The long length you may be thinking of is the post-acceleration
phase, where the packets are tweaked electrostatically or
magnetically to remove small velocity errors (this is needed to make
the mass catcher sufficiently small).
Paul