Regarding the metastability simulation described in W. Evans's article, how would you determine the time constant? For example, the time constant obtained between 1m to 2.713m would be quite different from that between 100m to 271.3m. Is there any rule of thumb for this or it's purely empirical?

Thank you!