Currently, steady-state probabilities are computed with the same method that is used for other linear equations. However, often power performs better for steady state but worse for reach probabilities. Hence, we should provide a second setting. This was originally mentioned in #49 .
