The purpose of 'stitching' in this instance is to increase the resolution of the final scan. I'll detail a scenario by way of illustration.
I have 6x7 negs. I load one into an enlarger neg carrier that I have, and place that flat on a Lightbox underneath my mounted Dslr which is positioned on a copy stand. The Dslr is fitted with a Nikon Af-D 105mm f2.8 Micro-Nikkor lens. That lens can focus to 1:1 ratio. Using a straight edge affixed to the Lightbox, I can slide the neg carrier across the field of view of my Dslr in such a way that I can easily shoot one half of the 6x7 neg, slide it over, and then shoot the other half. The camera remains stationary, as does the light source. Indeed with a little more effort I could easily shoot the 6x7 neg in quarters. The reason to do this is to double, or even quadruple the resolution of the scanned file, when the various exposures are stitched together in software.
Thus far I've limited my 6x7 scans to two exposures - each of which is on a 36 megapixel full frame sensor. Given my negs are 400 ISO, I'm getting sharp grain across the field of view. Combined files produce good, useful scans. I'd like to try scanning 4x5, using a similar set-up. Files need to be stitched, and for 4x5 that might become more cumbersome given I'd like to retain 4x5 resolution (within the limits of the system etc). Again, my question is about software options / workflows for stitching these files rather than going over hardware set-ups etc. Thanks for your help.