Less theatre, more evidence

I am becoming less impressed by demonstrations and more interested in examples where a system works repeatedly under real constraints.