I was going to post that this would have been cool a year ago, but now that MongoDB has interchangeable backends there's not much to gain from a high level project like this.
But then I saw that this project started two years ago, and the last commit was a year ago.
From my experience of MySQL's pluggable engines, I learned that this actually complicates things: the user interface ends up being awkward, optimization is harder, adding features is made much harder.
Investigating standalone alternatives therefore seems like a good thing, even for projects starting today.
But then I saw that this project started two years ago, and the last commit was a year ago.