I'm thinking that you're right: if the placebo effect reduces the error rate by 50%, then the environment definitely needs a permanently installed placebo like these vests.
All I'm saying is the nurse example doesn't really prove the main argument, which is that fewer interruptions equals fewer errors. Right now, it appears everyone reading this wants to apply this seemingly proven technique in their own work place, when (maybe) all they need is a device that reminds them that 1.) they make mistakes and 2.) their mistakes are being measured and observed.
One day, will someone notice that the vests have no obvious link to the tasks and promote removing hem for cost cutting?