I agree, Mary, I think they're the same person too, and that the death recorded in the Alabama state record was of an 87 year old person, in spite of what the online index now says. We can't see the original record to compare but any of the following errors could have happened: the informant could have been mistaken, the coroner misunderstood the informant's accurate information and wrote 97, the coroner wrote 87 but it looks like 97, the indexer saw 87 but typed 97, or the indexer genuinely thought it said 97, or or or.
The parents, spouse, and locations matching is strong evidence that these records are for the same individual. And the research/conflict note on the profile about the AL death record will help future researchers understand why duplicate profiles/dates of birth for this ancestor may tend to proliferate in online trees.