Good question. In General I agree with using "negative" sources as strong evidence, until my research either provides a burden of proof that has ruled out all other possibilities, or a record is found. I would use Research Notes as a way to track this.
I have encountered situations where people have been missing on the census and spouse is marked as Widowed. Only to discover later that the "spouse" was actually the other partner, gender incorrectly recorded by the enumerator of the census. Back then women used their husbands first name....Mrs John Doe and the enumerator doesn't record the Mrs. etc.
I have also found "missing" people that were totally passed over in the transcription of the original document. They won't come up in the search engine of the census, you might think they are dead, but if you search page by page you might find them.
They also seem to "disappear" when their name is recorded/transcribed with such a huge difference in spelling it appears to be a totally different surname, this is the worst one and takes tedious inch by inch searching to rule it out.
They might be visiting a relatives home or even friends, possibly in another country. I have found my relatives visiting in the US, or I find a border crossing/immigration/ships list record to explain things.
One time I even found a missing relative (for another genealogist) with a death recorded in another Country. I looked there because I discovered they had extended family in that Country. Totally surprised the PM!
That said, the addition of a source that just goes to a record set should have been added into Research Notes with an explanation about what search was done and the results of that search. You are really looking at a source that by the absence of an explanation is relatively useless and should be put in research notes as a follow up or removed possibly. Unless you undertake to verify the result. IMO