That seems wrong.
I've reconstructed the order of events using the timelines on all the posts in question:
01:23:35Z: user14111 posted a comment (now deleted) on the newer question which gave the title of the answer.
01:26:44Z: The user who asked the newer question (Arlen) posted a comment saying:
User14111 you are brilliant! Thank you!
This comment can be construed as acceptance according to our policy on closing story ID questions.
01:34:55Z: user14111 posted an answer to the newer question. The title is the same as given in the initial comment.
01:47:51Z: user14111 cast a binding close vote on the older question.
01:59:38Z: Arlen officially accepted user14111's answer on the newer question.
02:03:29Z: You posted this meta question. Since this occurred only a few minutes after Arlen accepted user14111's answer, it makes sense that you thought the newer question didn't have an accepted answer.
The order of events is difficult to follow, and it's completely understandable that you think it "seems wrong". However, the closing of one question as a duplicate of the other was technically in accordance with our policy on story ID duplicates.
Having established that the two questions can indeed be closed as duplicates, the issue becomes whether to close the newer or older one as the duplicate of the other. Our policy is to leave the better of the two duplicate questions open and close the other one, even if the older one ends up closed. Both the question and answer on the newer one are more descriptive, so I believe it is the right choice to close the older one as a duplicate of the newer one.