Analytical validation for sequencing is not one experiment. It is five parameters, each with its own sample requirement, and the reference material is usually the constraint.
The break-even point is not a sample count. It is a sample count at a given turnaround requirement, and the turnaround requirement is what usually decides it.