Do you mean that you saw actual differences when using Javascript in a normal website? Or that they were slower than the alternatives in a synthetic benchmark?

Do you mean that you saw actual differences when using Javascript in a normal website? Or that they were slower than the alternatives in a synthetic benchmark? It would be really surprising that those were the performance bottlenecks.