Using image data other than the filename would have performance implications. Pairing by filename should be nearly instantaneous, anything else would require reading significantly more of the file.
There are a few ways to avoid this in the future:
1) In the camera, change the filename prefix to something unique per camera, so even if the numbering overlaps, the filenames won't.
2) Ingest to different subfolders of the primary destination. You could do this using the Ingest Sequence variable to create a separate folder per ingest, or even set up something like a code replacement that uses the camera's serial number to then name the folder accordingly.