How can I increase the accuracy of Goal-seek in calc?

I am guessing that calc normally uses double-precision floating point, or even long doubles, but the results of goal-seek tests seem to stop at about single precision. Is there a control that sets the accuracy?

Goal Seek is a basic tool using a Regula Falsi approach with a limit of 100 iterations and a delta of 1E-6. You’ll have more options and control with the Tools → Solver… engines.