That certainly sounds plausible. I do wonder if these overly purple images are basically what the developed negatives look like, and that the lab had corrected the initial scans with software, but then failed to do so the second time around. I've emailed them about what might have happened, but I might also try run these scans through something like Darkroom.