These microbenchmarks don't really do it for me. It would be much more interesting to see a real test application being benchmarked. Something with users, objects, pages that have content.
It would be more fair to look at the code, complexity and performance drops when you actually build something more real.
Also, I know Erlang is awesome and super fast. But personally I would rather spend the money on a good horizontally scalable architecture and extra hardware if that means we can build apps in a more mainstream language like for example Python or Ruby.
These microbenchmarks don't really do it for me. It would be much more interesting to see a real test application being benchmarked. Something with users, objects, pages that have content.
It would be more fair to look at the code, complexity and performance drops when you actually build something more real.
Also, I know Erlang is awesome and super fast. But personally I would rather spend the money on a good horizontally scalable architecture and extra hardware if that means we can build apps in a more mainstream language like for example Python or Ruby.