Well that would depend exactly what we're talking about. "Cognitive capabilities", wasn't it? Maybe exclude babies but include drunks, if you like. I don't know who really belongs on the list, which is kind of the point.

Category error started at "we should not treat humans with impaired cognitive abilities as equals".

Yes, there's a good chance that whatever combination of metrics gets accepted to draw the consciousness/moral weight boundary in the future, it'll have some overlaps like in that infamous quip about "smartest bears" and "dumbest tourists". Applied to a spot measurement, it'll likely put some subset of people - drunk, cognitively impaired, sleeping, children under N months, etc. - below the "conscious/sentient" threshold, and below some animals.

Which is why such criteria only make sense at population lifetime level, not individual level. That's both aggregate and temporal ranges. Temporal, because children grow up, drunk people sober, impaired people may get better and may have been better in the past. Aggregate, because once you go down to individual level, you invite most horrible crimes ever committed.