As I stated in that other thread, there are few records of deaths for New Netherland, but there's no shortage of unsourced death dates from people's Internet genealogies. Accordingly, I wouldn't get hung up on pinning down a death date. Since he seems to be the only Roelof Kierstede in Kingston until his grandchildren of the same name grew up, I would record the records where he witnessed baptisms in Kingston in the Church Records subsection of his profile and use them to bracket a possible death date.
It's not uncommon for records for different children of the same family to show up in different New Netherland churches, even if the family didn't move. Each of the handful of churches in New Netherland apparently served a fairly large geographic area. Baptisms were performed by traveling domines (ministers) and were recorded at the church where the domine was based. We can't tell where a family was living based solely on where their children's baptisms were recorded. Regardless of that, that Albany baptism in December 1685 is mysterious. It's not at all obvious that it was the child of this man. Not only is the geography different, but the name is also different. The record says "1685. Dec. 13. Roeloff (bo. after his father’s death), of Roelof Kersten. Wit.: Cornelis Gysbertsz. By Anna Van Schayk." That name could be Roelof Kiersted, but it also easily could be (for example) a man named Roelof Karstens (where the last name is a patronymic for the son of Karsten). Analysis of witnesses might help identify the family. The child Aldert baptized in Kingston in November 1685 is clearly identified with this man and his wife, so the child named Roeloff must be someone else's child. The profile for that child named Roeloff is currently connected to this man as father for want of a better connection, but since there seems to be no further record for him, that connection isn't doing any harm, but the connection should be marked uncertain while the situation is researched.
It is tempting to build profiles for entire families based on old genealogical sources like that 1882 article in NYGBR, but those old genealogists do steer us wrong sometimes, and chances are good that there is other material from sources such as translated Dutch records that weren't available to 19th-century researchers. (Even the transcriptions of the various church records hadn't yet been published in 1882.) And we definitely cannot rely on those sources for transcriptions of baptism and marriage records -- there's a long history of published genealogies reporting the author's interpretation of a person's name, rather than the name that actually appears in the record (which in some New Netherland families might be very different from the name in the published genealogy).
The transcribed Albany church records are readily available online from a number of sources, so you can check them easily. I like the website https://mathcs.clarku.edu/~djoyce/gen/albany/refchurch.html , but the best sources are the original Holland Society of New York books (also online, for free) from which the content on that website was derived.
