Not really. Scala.reflect isn't the thing that'll be the thing anymore. Scala.meta is, and it's not considered done or stable yet.
(At least last I heard.)
Say what you will about the current extractor, but in 95% of cases it works, and in another 1% it's relatively straightforward to make it work, and that's with no development effort in some time. Those are some pretty big votes in its favor.
We could probably develop a parallel scala.meta-based extractor for optional use though.
Will post some ruminations on the list when I get a chance.