No doubt about its usefulness in areas with limited, or no, radar coverage.
But, I doubt that offset tracks would do much good there. Where the terrain
environment would permit parallel tracks, the publication of parallel tracks
as separate database routes would probably be much safer than issuing
parallel offset tracks in a non-radar environment.

The push for lateral offsets is greatest for non-radar environments, where
controllers are less likely to catch an operational error that puts two planes
on the same track and altitude. As I stated in an earlier post, one of the
proposed criteria is that the use of lateral offsets be transparent to the
controller. So offsets will probably be implemented as procedural changes
(for example, always fly 1 NM to the right of centerline if able), not as
clearances.