Certainly I have a lot of individuals where there is a Y in the death. I didn't put them there and I assume that in the quest for a stable program one of the programs I have used did so.
My copy of Reunion allows me to choose whether I have death as a default event (fact, but that is another debate), but my FTMM automatically has Died, but allows me to add a death fact. All my individuals have a death event therefore which may be populated or not.
The question is whether you should populate this in a way which will aid your research or whether you should do it in a way that conforms with the gedcom standard.
You then have a secondary question of whether your program outputs a gedcom that conforms with the standard, and even if it does, will the receiving program treat it correctly.
I recently spent several hours trying to validate a death given in a PMT as 1982. When I finally found the answer on the Internet, the individual had in fact died over two decades later and as it was her obituary that I found there was no doubt. This is not to say that whoever originally recorded the 1982 date had made an error. I suspect that originally it had been "after 1982" but in passing around the "after" had disappeared.
Your first duty I believe is to make entries in a form that will assist your research/suit any trees/reports you generate. I have therefore date wise entries such as between ......., before, after or blank. There may be a place even where there is no date, or there may not. I would not enter a Y as I do not believe this assists me and I do tend to delete existing ones.