A few years ago I wrote a paper with three collaborators, where we introduced a model to describe a specific phenomenon. In the end we were not fully happy with the result, so we wrote it up, put it on the arXiv without submitting to a journal, and agreed to revisit the subject after a small break.
Approximately a year later we gathered again, re-started almost from scratch, and came up with a better model, which resulted in a paper that we quickly submitted, and had accepted in a good journal.
A bit more than a year later, I now see that even though the second paper gets cited, the first still rakes up some amount of citations. We ourselves think that the second one is better, but obviously there are people who disagree.
What to do with the first paper? Should it be submitted to a journal to undergo peer review? And would it not be strange to try and submit the first paper to a journal, pretending that the second paper never happened?