Yep, and others say it’s less than 1%. Neither of them are lying, but their talking about different kinds of observations.

My assumption is scheduling should solve most of these issues in terms of large instruments.