The blog post assumed (and made that assumption clear) that the age was correct. Our discussion was based upon that premise. Genealogists always have to make assumptions before any analysis can begin--it's the acknowledgement of those assumptions that is important. The point of that post was not about the assumption; it was about the analysis if that assumption were true.
The question was raised though that the age could have been rounded to the nearest month and, if that age was rounded, was it rounded up or rounded down? The thought that the age was rounded in some way is a legitimate one. The question is "how would the rounding have been done?"
Up or Down?
If the informant rounded up, they would have indicated Belinda was 66 years and 6 months if she was into the 5th month of her 66th year.
If the informant rounded down, they would have indicated Belinda was 66 years and 6 months only if she was over 66 years and 6 months and less than 66 years and 7 months.
My concern with the rounding is that I cannot be certain how it was done and it requires me to "guess" at how the rounder rounded the date. I'm not even certain who the rounder was--was it the son who paid for the stone or was it the stonecutter?
My thought is that "if" the months were rounded, they were rounded down and not up. My rationale for this is that even if I am 45 years and 10 months of age, I am likely to say I'm 45 and not as likely to say that I'm 46 (although I may say "I'm going on 46.").
The other concern is how big of a deal is this rounding? It only makes the month of birth one month different. The difference is likely to be significant in only rare cases where the researcher is trying to sort out two people with similar names born extremely close to one another.
My rule of thumb is to go with what the document says and indicate that the analysis is based upon that.