Judging from slide mounts and color quality (and color loss), I am guessing that some of the slides I'm scanning were probably developed by the photographer (i.e., my grandfather) rather than a lab. The two photos in this post show some stark differences, and I believe they also represent the developing difference.
The photo showing a bus with a Sony ad appears to have been developed in a lab, and (after automatic color restoration) it has reasonably good color and it's not backwards. The photo of the other street scene has (after automatic color restoration) very washed out color (the uncorrected film has a distinct red tint, apparently from aging), and text within the image shows that the image is backwards.
However, this leaves me wondering about the (home-developed?) photo being backwards. The scanner instructions are clear in that the emulsion side should be up and the base side should be down when loading slides. The "Sony bus" photo shows that this comes out correctly from the (lab-developed?) slide. Yet the other street scene was also scanned with the emulsion side up and the (shinier) base side down. Development alone wouldn't make the emulsion change sides! So, why is the image backwards?
Two thoughts: First, the film might have been loaded into the camera wrong. Yet this doesn't explain why this problem seems to come up with slides that may have been developed at home but not those apparently developed by a lab. Second, there might be something unusual about the film itself. Again, this doesn't explain why the problem seems to coincide with slides that don't appear to have been developed by a lab.
Decades later, long after my grandfather passed away, there is no way to get further information about the film, the developing, or anything else. So many questions, so few clues!