It's good to see a lot of the techniques proven useful by the Lua interpreter showing up for other languages. The author states that he's no longer convinced of the importance of JITs. I still think that they're going to be an increasingly important advancement for dynamic language implementations. For example, the LuaJIT developed by Mike Pall shows speedups of 3-5x over the (already very fast) reference Lua interpreter.
http://luajit.org/luajit_performance.html
Thank goodness for open source. The best ideas and techniques will propagate until we have multiple high quality implementations readily available.
¡Viva La Revolución!