I don't really see that statistic as a problem. It seems to me that shipping slightly more content than most players are expected to finish is actually a good

I don't really see that statistic as a problem. It seems to me that shipping slightly more content than most players are expected to finish is actually a good thing, especially in traditional game development where you are selling a more-or-less complete product (where the business plan doesn't involve skimping on the core product to explicitly to sell expansion packs and other add-ons). It means there's enough content for most of the audience. There's very little correlation between content completion percentage and overall game enjoyment / purchase satisfaction.