In case you skipped the article it contains this nugget:
Take this example containing functional composition and higher-order functions:
reduce(operator.add, 0, map(function(x) return x^2 end, range(n)))
LUA converts this into 10 lines of assembler (2 jumps). Also, the LUA compiler is directly generating SSE.
Anyone have any good benchmarks comparing simple functional examples in LUA with Clojure or Common Lisp?