I'm pretty sure Haskell has raised the bar for level of abstraction. This is at least partially because it is of interest to researchers pushing the limits of what programming languages can do. Just an example:

http://math.andrej.com/2007/09/28/seemingly-impossible-funct...