I just retweeted Michael's tweet link this blog entry - this is really a very good classification of Lisp systems! It could be a potential time sink however, trying more implementations :-)