A superficial Data Analysis of Accepted answers
I wrote a couple of queries on Data.SE
This one shows the average rep of the answers with accepts vs the average rep against not accpeted
AVG Accept Rep
AcceptFlag acceptrep
------------ ---------
Accepted 2,100
Not Accepted 14,003
I was surprised by this result (are very high users' answers skewing things?) so I wrote this one.
Rep Breakdown of unaccepted answers
This separates posts by the asker's rep into groups of 10 and then finds the maximum rep in each group.
decile rep
----------- ------
1 29
2 181
3 569
4 1,204
5 2,212
6 3,955
7 7,259
8 13,836
9 34,158
10 336,525 <-[Well we know who asked that question]
I'm not a statistician but it would seems that there are plenty of questions that aren't accepted are by high rep users.
This may be because after awhile new users will go back and accept answers once they learn about it while (as @awoodland comments) while high rep users will ask hard questions that don't get answered because they're too hard. Also rep isn't static, overtime rep accrues, but the acceptance/non-acceptance often is after a certain point.
Honestly though since I don't break the 200 daily rep limit very often its pretty rare that I actually care since its much easier to just get 2 upvotes.
Also very few badges are associated with answer acceptance 1 gold, 3 silver, and 1 bronze while there are dozens associated with score, not including all the tag based badges.