A few years ago I had published a mathematics journal article where for one of the results I had imposed extremely strong assumptions to ensure that given conclusion is true. Figuratively speaking, I now realise that I have used a hydraulic press to crack a peanut. Indeed, the same conclusion can be obtained with minimal assumptions.
Needless to say, I am deeply ashamed by this sloppiness, but I am not sure if this merits publishing an erratum. One co-author strongly opposes publishing an erratum, saying that we should simply add a remark to a follow-up paper we are writing, while the other couldn't care less. As the original result isn't strictly incorrect, I'm not sure what the appropriate approach is.