I really don't see the difference. If they can detect all the geometrically changed partitions, when they do the image, the existing unchanged partitions will be imaged normally and the changed partitions will be fully imaged. The resultant image will have all the unchanged partitions as well as any additional partitions that may have changed... same effort, a bit more efficient (unchanged partitions imaged only for changes, changed partitions imaged in FULL instead of all in FULL).
Seems safe to me... just a slightly different approach
Seems safe to me... just a slightly different approach