The ED₁₀ is more directly and consistently related to the observed dose-response data than the  regulatory potency measure (q₁*) based on the linearized multistage model.  In addition, the multistage and linearized multistage models are limited in their ability to reflect the observed dose-response data especially at lower doses -- because of the limited flexibility imposed by their structure (functional form) and regulatory policies that assume non-threshold behavior and force the models parameter values to be greater than or equal to zero regardless of the data.  Furthermore, the linearized multistage model has been designed for regulatory purposes to be biased with a strong tendency to overstate risks.  Therefore, my recommendation is to use the ED₁₀ as the basis for the dose-response characterization and not the regulatory potency measure (q₁*) based on the linearized multistage model.